On Tue, Mar 20, 2018 at 7:29 AM, Linus Torvalds
[off-list ref] wrote:

On Mon, Mar 19, 2018 at 2:43 AM, David Laight [off-list ref] wrote:

quoted

Is it necessary to have the full checks for old versions of gcc?

Even -Wvla could be predicated on very recent gcc - since we aren't
worried about whether gcc decides to generate a vla, but whether
the source requests one.

You are correct. We could just ignore the issue with old gcc versions,
and disable -Wvla rather than worry about it.

This version might also be an option:

diff --git a/Makefile b/Makefile
index 37fc475a2b92..49dd9f0fb76c 100644
--- a/Makefile
+++ b/Makefile

@@ -687,7 +687,8 @@ KBUILD_CFLAGS += $(call cc-option,-fno-reorder-blocks,) \
 endif

 ifneq ($(CONFIG_FRAME_WARN),0)
-KBUILD_CFLAGS += $(call cc-option,-Wframe-larger-than=${CONFIG_FRAME_WARN})
+KBUILD_CFLAGS += $(call cc-option,-Wstack-usage=${CONFIG_FRAME_WARN}, \
+               -$(call cc-option,-Wframe-larger-than=${CONFIG_FRAME_WARN}))
 endif

 # This selects the stack protector compiler flag. Testing it is delayed

Wiht -Wstack-usage=, we should get a similar warning to -Wvla for frames that
contain real VLAs, but not when there is a VLA that ends up being a compile-time
constant size in the end. Wstack-usage was introduced in gcc-4.7, so
on older versions
it turns back into Wframe-larger-than=.

An example output would be

security/integrity/ima/ima_crypto.c: In function 'ima_calc_buffer_hash':
security/integrity/ima/ima_crypto.c:616:5: error: stack usage might be
unbounded [-Werror=stack-usage=]

        Arnd