Added the 'no_randomize_layout' attr to clang

[Nixoncole]

Closes #6

Please ignore my poorly named git branch...

Outside of that, I think this should work. Next is docs, which will touch both of these files again.

This doesnt have to be merged if you dont want to @connorkuehl but here ya go

[Nixoncole]

Ah I'll be honest I did not know there was formalized acceptance criteria. I'll look around and see about emitting compiler warnings etc.

[tim-pugh]

Let's add this PR to a develop branch (a different one then the current one named develop :-) ) so we can prep for the RFC, but besides that I'll have to take a look at this tomorrow/when the rest of the bits come on in.

[connorkuehl]

Let's add this PR to a develop branch (a different one then the current one named develop :-) )

I might be mistaken, but I thought we were branching off only for features that would be merged AFTER we instate the code freeze. Is that correct?

[tim-pugh]

Let's add this PR to a develop branch (a different one then the current one named develop :-) )

I might be mistaken, but I thought we were branching off only for features that would be merged AFTER we instate the code freeze. Is that correct?

We've talked about this last Tueday

"We'll call a code freeze, send rfc, have a develop branch for the stretch, fix the feedback, and can resubmit again"

If a stretch goals is all good then we'll just do a pr from that branch (or branches) into our mainline before the code freeze. That is fine.

[Nixoncole]

Just added code that implements first attempt at emitting a compiler warning when a struct is tagged with both attributes "randomize_layout" and "no_randomize_layout".

Building now for test

[Nixoncole]

Just to let everyone know what the state of this is, in case they want to help...

This currently builds, and our only implementation additions are supposedly barebones for emitting a compiler warning. The issue is that when trying to compile a file, I receive...

NixonCole@Coles-MacBook-Pro-2 ~/clangRandStruct $ llvm-project/build/bin/clang poc.c -o poc
poc.c:2:10: fatal error: 'stdio.h' file not found
#include <stdio.h>
         ^~~~~~~~~
1 error generated.

When I compile with my system's clang (Mac OS), it compiles without issue. This leads me to believe that I have broken something somehow with my implementation.

[connorkuehl]

poc.c:2:10: fatal error: 'stdio.h' file not found This leads me to believe that I have broken something somehow with my implementation.

This doesn't strike me as a regression in your code. It looks like the compiler can't find your standard library. I don't know how to verify where or how or if it's properly installed on Mac. Try running your compile command with -v (ex: path/to/clang -v poc.c. You'll want to look for parts in the text where it says something like:

#include "..." search starts here:
#include <...> search starts here:
 /usr/local/include
 /home/kuehlcon/llvm-project/build/lib/clang/9.0.0/include
 /usr/include/x86_64-linux-gnu
 /usr/include

I checked out your branch and tried it out for myself.

Here is the version of poc.c:

#include <stdio.h>

struct mystruct {
    char *first;
    char *second;
}__attribute__((no_randomize_layout)) __attribute__((randomize_layout));

int main(void)
{
    struct mystruct m;
    m.first = "I'm the first!";
    m.second = "and I'm the second!";

    printf("%s\n%s\n", *(&m), *(&m + sizeof(char *)));
    return 0;
}

Here's the compile:

/home/kuehlcon/llvm-project/build/bin/clang -g -Wno-format poc.c -o rand
poc.c:3:8: warning: struct declared with 'randomize_layout' and 'no_randomize_layout'
      attributes. attribute 'no_randomize_layout' takes precedence
struct mystruct {
       ^
1 warning generated.

Checking structure layout with pahole shows no difference :+1:

As well as a quick run to make sure nothing's happened:

kuehlcon@randstruct-dev:~/randstruct-testing/poc$ ./rand
I'm the first!
and I'm the second!
kuehlcon@randstruct-dev:~/randstruct-testing/poc$ ./reg
I'm the first!
and I'm the second!

Nice work! I'll go through and update my review.

[connorkuehl]

I have no idea why Github is displaying all these commits in this PR; they've already landed in develop.

It's easier to review the PR by seeing the branch diff here.

[Nixoncole]

This looks pretty good if you ask me! Thank you for providing the link to an easier view.