ipv4/tcp.c:4234:1: error: the frame size of 1152 bytes is larger than 1024 bytes [-Werror=frame-larger-than=]

[Arnd Bergmann]

On Fri, Sep 17, 2021 at 8:16 AM Brendan Higgins
<brendanhiggins at google.com> wrote:
>
> On Thu, Sep 16, 2021 at 10:39 PM Brendan Higgins
> <brendanhiggins at google.com> wrote:
> >
> > On Tue, Sep 14, 2021 at 3:04 PM Arnd Bergmann <arnd at arndb.de> wrote:
> > >
> > > On Tue, Sep 14, 2021 at 10:48 PM Brendan Higgins
> > > <brendanhiggins at google.com> wrote:
> > > >
> > > > On Mon, Sep 13, 2021 at 1:55 PM Shuah Khan <skhan at linuxfoundation.org> wrote:
> > > > >
> > > > > On 9/8/21 3:24 PM, Brendan Higgins wrote:
> [...]
> > Alright, I incorporated all the above into a patchset that I think is
> > ready to send out, but I had a couple of issues with the above
> > suggestions:

Thanks a lot for those suggestions.

> > - I could not find a config which causes a stacksize warning for
> > sdhci-of-aspeed.

I keep a history of my randconfig builds. This one only happened
once before I fixed it, it may depend on some other combination of
options. See my original defconfig file at
https://pastebin.com/raw/XJxjVGYa
rand/0xAB2DD5A0-failure:/git/arm-soc/drivers/mmc/host/sdhci-of-aspeed-test.c:47:1:
error: the frame size of 1152 bytes is larger than 1024 bytes
[-Werror=frame-larger-than=]

> > - test_scanf is not a KUnit test.

I have three defconfigs for this one, all on x86-64:

rand86/0x30AD57FB-failure:/git/arm-soc/lib/test_scanf.c: In function
'numbers_list_field_width_val_width':
rand86/0x30AD57FB-failure:/git/arm-soc/lib/test_scanf.c:530:1: error:
the frame size of 2488 bytes is larger than 2048 bytes
[-Werror=frame-larger-than=]
rand86/0x30AD57FB-failure:/git/arm-soc/lib/test_scanf.c: In function
'numbers_list_field_width_typemax':
rand86/0x30AD57FB-failure:/git/arm-soc/lib/test_scanf.c:488:1: error:
the frame size of 2968 bytes is larger than 2048 bytes
[-Werror=frame-larger-than=]
rand86/0x30AD57FB-failure:/git/arm-soc/lib/test_scanf.c: In function
'numbers_list':
rand86/0x30AD57FB-failure:/git/arm-soc/lib/test_scanf.c:437:1: error:
the frame size of 2488 bytes is larger than 2048 bytes
[-Werror=frame-larger-than=]

https://pastebin.com/raw/jUdY6d3G is the worst one of those

> > - Linus already fixed the thunderbolt test by breaking up the test cases.

Ok.

> > I am going to send out patches for the thunderbolt test and for the
> > sdhci-of-aspeed test for the sake of completeness, but I am not sure
> > if we should merge those two. I'll let y'all decide on the patch
> > review.
>
> Just in case I missed any interested parties on this thread, I posted
> my patches here:
>
> https://lore.kernel.org/linux-kselftest/20210917061104.2680133-1-brendanhiggins@google.com/T/#t

Thanks! I'll reply to the particular patch as well, but I don't think
that this is
sufficient here:

+CFLAGS_bitfield_kunit.o := $(call
cc-option,-Wframe-larger-than=10240) $(DISABLE_STRUCTLEAK_PLUGIN)

If 10KB is actually needed, this definitely overflows the 8KB stack on
32-bit architectures.

         Arnd