grsecurity forums

Does the stack randomization features of grsecurity screw with IBM's stack protector? (probably a stupid question, but I'd like to know)

flamingice wrote:Does the stack randomization features of grsecurity screw with IBM's stack protector? (probably a stupid question, but I'd like to know)

they should not (i don't see what feature of propolice would need knowledge of absolute addresses). in fact, now that OpenBSD has included propolice and they also have stack randomization, you can be fairly sure it works fine.

I've compiled several of my apps with propolice, and have had no problems with it and PaX.

-Brad

I know they work together, I just wasn't sure that propolice would be effective with stack randomization on. (since propolice needs to put stuff in certain places, blah blah, don't know much about this stuff)

flamingice wrote:I know they work together, I just wasn't sure that propolice would be effective with stack randomization on. (since propolice needs to put stuff in certain places, blah blah, don't know much about this stuff)

what propolice does to the stack is that it reorganizes its layout (that is, each function's stack frame), this does not rely on absolute addresses therefore the global stack randomization (think of it as shifting, something that you would also get if you defined an environment variable of a random length for every task) that PaX does won't interfere with it.