在 2025-7-3 20:06, Igor Kostenko 写道:
Fix stack protection bootstrapping issue where the canary initialization function itself triggers false positive stack overflow detection.Signed-off-by: Igor Kostenko <work.ker...@gmail.com> --- mingw-w64-crt/ssp/stack_chk_guard.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/mingw-w64-crt/ssp/stack_chk_guard.c b/mingw-w64-crt/ssp/stack_chk_guard.c index 3ff22e020..7e4a6c1a0 100644 --- a/mingw-w64-crt/ssp/stack_chk_guard.c +++ b/mingw-w64-crt/ssp/stack_chk_guard.c @@ -10,7 +10,11 @@void *__stack_chk_guard; -static void __cdecl __attribute__((__constructor__)) init(void)+// no_stack_protector attribute added bellow, +// to disable stack protection checks even before the canary is initialized
Would you please explain the exact meaning of 'canary'?
+// check will always fail at the end of this function +// initial value (0) != initialzed value (rand or dedbeaf const)
Here are two typos. While I understand the reason, this explanation doesn't seem so straightforward. What about: // This function requires `no_stack_protector` because it changes the // value of `__stack_chk_guard`, causing stack checks to fail before // returning from this function.
+static void __cdecl __attribute__((__constructor__, no_stack_protector))
init(void)
{
unsigned int ui;
if (__stack_chk_guard != 0)
For consistency reasons I'd prefer this be `__attribute__((__constructor__, __no_stack_protector__))`. This line is too long now, so maybe we can have
__attribute__((__constructor__, __no_stack_protector__)) static void __cdecl init(void)The `no_stack_protector` was added in GCC 11 and Clang 7, but both have a convention of ignoring unknown attributes, so it's probably fine to leave out a check.
-- Best regards, LIU Hao
OpenPGP_signature.asc
Description: OpenPGP digital signature
_______________________________________________ Mingw-w64-public mailing list Mingw-w64-public@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
