Your code should be taken out back, lined up against a wall, and machine-gunned

http://www.cygwin.com/ml/cygwin/2005-08/msg00520.html

by RyanMcGreal • 16 years ago

94 73 16 years ago

chacha102• 16 years ago

  Then the bleeding corpse should be hung, drawn and quartered.
  
  Then burnt.
  
  Then the smouldering rubble should be jumped up and down on.
  
  By a hippo.

By a hippo....

JeanPierre• 16 years ago

It is a natural choice. 3,600 lb is more than the weight of any human in the world, and elephants can't jump.

sliverstorm• 16 years ago

Well, you'd at least hope so.

afhof• 16 years ago

Why was he complaining about the stack pointer first? Isn't this line:

st1 = (void *)malloc(5000) + 5000;

completely invalid too? You can't (I think it's undefined) increment a void pointer.

tptacek14• 16 years ago

It may be undefined, but gcc accepts it without warning under -Wall and does exactly what you'd expect. Note that (a) casting malloc's return is evil, (b) casting malloc to void* is silly since it's declared returning void* already, and (c) casting malloc to void* so it can be received as a char* is also pretty goofy.

NickPollard• 16 years ago

This might be a stupid question, but what are you meant to do if not cast malloc's return?

My C is a bit rusty, but how else are you meant to allocate data of different types on the heap?

fintler• 16 years ago

Yep, it's basically undefined because a void type has no known size to do pointer arithmetic with.

koenigdavidmj2• 16 years ago

GCC whines and moans, but treats it like a pointer to char.

scdlbx• 16 years ago

Actually, since st1 is a char* and the sizeof char is 1, that is a valid declaration. It just probably isn't doing what the writer expects. It allocates 5000 bytes and then moves the pointer to point at the very last byte.

edit: Though free()ing it would not work so well until you move the pointer back to the beginning of the 5000 bytes.

azim• 16 years ago

I think that's what the writer expected. On x86 the stack starts at a high address and grows downward to a lower address.

mfukar3• 16 years ago

Yes, arithmetic on void * is disallowed. However, gcc accepts it as an extension: http://gcc.gnu.org/onlinedocs/gcc-2.95.3/gcc_4.html#SEC78

mattmight• 16 years ago

Or, the code could be released under the CRAPL:

http://matt.might.net/articles/crapl/

j_baker• 16 years ago

I'm a little bit slow with the low level stuff. Could someone explain what's so horrific about this code?

chipsy• 16 years ago

C's function call mechanism requires internal maintenance of a call stack, which the programmer ordinarily never has to touch. The broken code author claims that he needs to manipulate said stack to write his Scheme interpreter. He does it wrong in a way that might pass the trivial "hello" test case under Linux, and blames Cygwin when it doesn't even pass that.

ww52012• 16 years ago

For one thing, he's using assembly of one type of machine and expects it to work on different type of machine. He confuses Linux as a machine abstraction and demands that Cygwin running on Windows to be the same.

For the low level mess, he sets the stack frame pointer to a uninitialized heap-allocated buffer which contains God-know-what garbage. When main() returns, it pops its return address from the stack frame pointer, which contains garbage and crashes. The printf() might sometime work depending on how the compiler emits code to set up the stack frame.

It just shows how dangerous some ignorant programmers can be.

mrfu• 16 years ago

For further information http://insecure.org/stf/smashstack.html is an interesting read.

rntz• 16 years ago

Uh, actually, his original program (http://www.cygwin.com/ml/cygwin/2005-08/msg00542.html) did restore the stack pointer before returning from main. He "wanted to cut [the example's size] down for the sake of the mailing list", so he left that bit out.

tptacek14• 16 years ago

Say what? Nothing you've said here makes sense.

First, he's writing a nonportable language runtime. For Cywgin. It's not a fair critique to say "he's making the mistake of writing x86-specific code".

Second, you don't have to initialize a stack before you use it. The stack isn't magically cleared before it gets used. In fact, that's rather the point of having the stack implemented with a single pointer, and is also the reason that in C you have to explicitly initialize (say) integers before using them.

Third, there's nothing "dangerous" about what he's doing, in that what he's doing can't ever work. He created a temp stack, write code that attempts to switch into it, and never wrote the code to switch out of it. That's not going to slip into production. But that's also not a fundamental critique of the approach.

What he's trying to do works fine, including under Cygwin, when completely implemented. Despite much peanut gallery whining that it can't possibly work, or that this guy should give up on doing inline/threaded/coroutine stacks and instead implement some CS101 interpreter instead.

dedward• 16 years ago

gcc relies on the stack to handle function calls.

In this case he's using in-line assembly to modify the stack pointer.

That's bound to end in bad results.....

lallysingh• 16 years ago

The asm statement puts the pointer (the malloc(5000)+5000's result) into the stack pointer. When the arguments are passed into fprintf(), they're pushed on top of that frame (presumably back into the malloc'd block).

In my book, a worthy attempt to grow your stack on the heap, but you've really got to be ready to hit the asm debugger when things go awry.

Edit: as mentioned in other comments, the mistake was (at least) in not restoring the stack pointer before main returned.

amalcon• 16 years ago

The short answer is that he's doing low-level stack-twiddly stuff using inline assembler in C. It's one of the standard perils of mixing languages: the C compiler assumes (and attempts to ensure) that it's the only thing manipulating the stack pointer, and the inline assembler violates that assumption.

llaxsll2• 16 years ago

If you actually read further along in the thread, it looks like this guy is trying to implement coroutines. However, he cut some essential part of a sample code that he was initially looking at and then assumed that the problem is with cygwin. And then came some silly flame war.

I personally haven't done anything with coroutines but looks like from a low level perspective, ultimately what he is trying to do is not insane… aka do some form of task switching. Arguably, he shouldn't be doing this by himself and he should've resorted to something that is already available. Just a quick look at (http://en.wikipedia.org/wiki/Coroutine#Coroutine_alternative...) should've been sufficient. Yet.., maybe he had some reason to try and do it himself. Looks like at the end he managed to figure out what he wanted to do.

xpaulbettsx2• 16 years ago

While I always enjoy a good rant, I don't see why it had to be so vicious and insulting. Yeah, the guy's clearly pretty confused (if he wanted to do these kind of tricks, I'd try to use setjmp, fiddle with the structure, then longjmp to it), but there's no reason to be mean about it.

ww52012• 16 years ago

I think it's because his indignant demand of the Cygwin team to fix this "bug" and make it work for him. Basically it's the attitude that I know my way is the right way, now fix the platform so that my program runs.

xpaulbettsx2• 16 years ago

I don't see any demands in the original mail:

http://www.cygwin.com/ml/cygwin/2005-08/msg00504.html

He questions whether it "should be fixed" in Cygwin, a perfectly reasonable request, and he fully admits that he isn't sure whether it's a problem in Cygwin or not.

silentbicycle2• 16 years ago

Yeah, but blunt brevity and "go read X instead, you're going about it wrong" would have been better for everyone involved.

mfukar3• 16 years ago

Because he had the same attitude with this guy: http://bugs.php.net/52435

llaxsll2• 16 years ago

:).

From the comments: I get this problem too. It only seems to happen if your IQ is less than 70, though.

Lazlo_Nibble• 16 years ago

While I always enjoy a good rant, I don't see why it had to be so vicious and insulting.

The cygwin maintainers are not known for their patience and affability.

tptacek14• 16 years ago

While setjmp/longjmp is a good trick (it's how Hanson did his thread library in CII), it's actually an even less good approach by the standards of this mailing list; there are even fewer constraints on how Cygwin needs to implement longjmp than there are on how it manages %%esp.

jsn• 16 years ago

What are the problems with setjmp/longjmp? (a honest question; I've drafted a proof-of-concept implementation of C coroutines yesterday, just to see how long it will take me to get it working -- http://gist.github.com/491302).

boredguy8• 16 years ago

Best part is in the follow-up discussion:

  > Do you actually realise that not all computer languages can 
  > be implemented with a model of one stack?
  > 

  Do you actually realize that there are far better ways to implement stacks
  for interpreters than dinking with esp?  Ways that actually have a chance of
  working?  I suggest you investigate those alternatives instead of trying to
  do something that cannot work, and then getting all pissy when people try to
  dissuade you from wasting your time.

  And don't make me break out my Turing Machine on you about that statement of
  yours.

tptacek14• 16 years ago

Strong disagree. Korn's response is at least funny. The rest of the responses are petulant, somewhat ill-informed, and orthodoxical. The argument over whether C does or doesn't have a runtime is particularly tedious. You get the impression that these people would all be very, very upset at the Detours paper, except that Detours works so well they'd know a priori they'd lose the argumet.

The original commenter --- despite his very fragile code --- is right: for what he wants to do, he needs to create temporary stacks. What he doesn't have is a mechanism to allocate and swap in and out of those stacks (he's hoping that switching stacks is a 2-liner that just relies on knowing the inline asm incantation for changing ESP).

Regardless, it is a uniquely and flagrantly bad way to educate (both the questioner and the world who's reading the thread) to start from the premise that "you're not ready to build the thing you want to build" --- at least when "the thing" we're talking about is systems code. 90% of C programmers are not "ready" to build production C code that will get deployed in hostile environments, and as a result, in every web project we ever do, we have to embark on a tedious inventory of all the 3rd party stuff our clients use to find the horrible C ZIP library they accidentally included. No C coder is blameless.

Unless I missed a whole leg of this thread (the part where they explained temp stacks and stack switching to this guy, who, in challenging the notion that C has a real runtime or the notion that the OS should [heh] be controlling ESP, is demonstrating more on-the-ground systems programming knowhow than several of the commenters), I think this is a remarkably douchey example of a programming help thread.

ww52012• 16 years ago

Why don't you show us a working version of the code instead of spewing endless argument?

ivenkys• 16 years ago

I for one would pay good money to get a response like this.

He points out the errors,the incorrect assumptions and offers pointers of what to do. The general standard of discourse on most mailing lists tend to be quite a bit lower than this.

nailer• 16 years ago

There's a great story about a heckler at a Windows NT conference:

MS speaker: and we've got POSIX, and we've got POSIX ACLs, and we've got a Unix Subsystem and we've got Korn Shell, and it works just the same way as under Unix.

Heckler: NO IT DOESN'T

MS speaker: er, ok. It does work. Anyway...

Heckler: NO IT DOESN'T WORK

MS speaker: What doesn't work?

Heckler: (rattles off differences in behavior and brokenness of ksh on NT)

MS speaker: we'll we've had our top guys working on this ...

Heckler: I'M DAVID KORN.

Room erupts in laughter.

Update: source: http://slashdot.org/articles/01/02/06/2030205.shtml. See question 5.

tokipin• 16 years ago

we realllly need people like that at political speeches

edge17• 16 years ago

Is that the same Dave Korn that did Korn Shell?

Confusion4• 16 years ago

In case you haven't gathered from other responses in this thread or are only reading this in your 'threads' view: yes.

astrange• 16 years ago

Er, no he isn't.

(The older David Korn actually wrote a sort-of-competitor to Cygwin, called UWIN.)

thibaut_barrere• 16 years ago

Recruiting tip: don't hire someone who is able to answer this way to someone, no matter how skilled.

Your code should be taken out back, lined up against a wall, and machine-gunned

Related Stories