[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20161101223025.GG5749@port70.net>
Date: Tue, 1 Nov 2016 23:30:25 +0100
From: Szabolcs Nagy <nsz@...t70.net>
To: "LeMay, Michael" <michael.lemay@...el.com>
Cc: "musl@...ts.openwall.com" <musl@...ts.openwall.com>
Subject: Re: [RFC PATCH v2 1/4] disable check for buggy brk
implementations when SafeStack is enabled
* LeMay, Michael <michael.lemay@...el.com> [2016-10-28 19:56:18 +0000]:
> The check relies on comparing the addresses of stack-allocated objects
> to addresses returned by the brk syscall. SafeStack moves the
> allocations to the unsafe stack, breaking the check. This patch
> disables the check when SafeStack is enabled.
>
> Signed-off-by: Michael LeMay <michael.lemay@...el.com>
> ---
> src/malloc/expand_heap.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/src/malloc/expand_heap.c b/src/malloc/expand_heap.c
> index d8c0be7..af51451 100644
> --- a/src/malloc/expand_heap.c
> +++ b/src/malloc/expand_heap.c
> @@ -13,6 +13,7 @@
>
> static int traverses_stack_p(uintptr_t old, uintptr_t new)
> {
> +#if !defined(__has_feature) || !__has_feature(safe_stack)
preprocessing this fails on any released version of gcc
https://godbolt.org/g/hHWEzI
if __has_feature is not defined it expands to 0 and 0(safe_stack)
is invalid syntax.
e.g. this works:
#ifdef __has_feature
#if __has_feature(safe_stack)
#define HAS_SAFE_STACK 1
#endif
#endif
> const uintptr_t len = 8<<20;
> uintptr_t a, b;
>
> @@ -23,6 +24,7 @@ static int traverses_stack_p(uintptr_t old, uintptr_t new)
> b = (uintptr_t)&b;
> a = b > len ? b-len : 0;
> if (new>a && old<b) return 1;
> +#endif
>
> return 0;
> }
> --
> 2.7.4
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
