linux-kernel - Re: [RFC 2/2] wl3501_cs: reduce stack size for KASAN

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAK8P3a2jHkrE57jYWuXOkyUSeFxy0jdT-pnL4GXP7z0bhHiTKQ@mail.gmail.com>
Date:   Tue, 25 Jul 2017 16:50:47 +0200
From:   Arnd Bergmann <arnd@...db.de>
To:     Kalle Valo <kvalo@...eaurora.org>
Cc:     "the arch/x86 maintainers" <x86@...nel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>,
        "H . Peter Anvin" <hpa@...or.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        linux-wireless <linux-wireless@...r.kernel.org>
Subject: Re: [RFC 2/2] wl3501_cs: reduce stack size for KASAN

On Tue, Jul 25, 2017 at 2:52 PM, Kalle Valo <kvalo@...eaurora.org> wrote:
> Arnd Bergmann <arnd@...db.de> writes:
>
>> Inlining functions with local variables can lead to excessive stack usage
>> with KASAN after a previous patch that modifies the outsb/insb helpers
>> on x86.
>>
>> drivers/net/wireless/wl3501_cs.c: In function 'wl3501_rx_interrupt':
>> drivers/net/wireless/wl3501_cs.c:1103:1: error: the frame size of 2232 bytes is larger than 1536 bytes [-Werror=frame-larger-than=]
>>
>> Marking the two callers of insb/outb 'noinline' prevents the compiler
>> from adding up the stack usage for each of the local variables passed
>> into those, reducing the maximum stack frame size to 800 bytes with
>> KASAN again.
>>
>> Signed-off-by: Arnd Bergmann <arnd@...db.de>
>
> Arnd, based on the discussion I'm dropping. Please let me know if I
> should take this still.

Thanks, that's good. The problem has become unreproducible and
I assume it's gone for good with the new x86 fix. In the unlikely
case some form of the problem comes back in another randconfig,
I'll post a new patch.

       Arnd