linux-kernel - Re: [syzbot] KASAN: stack-out-of-bounds Read in profile

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

Message-ID: <527ad07e-eec2-a211-03e7-afafe5196100@linux.intel.com>
Date:   Wed, 2 Jun 2021 16:35:11 -0700
From:   Andi Kleen <ak@...ux.intel.com>
To:     Josh Poimboeuf <jpoimboe@...hat.com>,
        syzbot <syzbot+84fe685c02cd112a2ac3@...kaller.appspotmail.com>
Cc:     bp@...en8.de, hpa@...or.com, inglorion@...gle.com,
        linux-kernel@...r.kernel.org, mingo@...hat.com,
        syzkaller-bugs@...glegroups.com, tglx@...utronix.de,
        x86@...nel.org, Peter Zijlstra <peterz@...radead.org>,
        Andy Lutomirski <luto@...nel.org>
Subject: Re: [syzbot] KASAN: stack-out-of-bounds Read in profile_pc


> profile_pc() assumes the return address is either directly at regs->sp,
> or one word adjacent to it due to saved flags, both of which are just
> completely wrong.  This code has probably never worked with ORC, and
> nobody noticed apparently.

I presume it used to work because the lock functions were really simple, 
but that's not true anymore.

>
> We could just use ORC to unwind to the next frame.  Though, isn't
> /proc/profile redundant, compared to all the more sophisticated options
> nowadays?  Is there still a distinct use case for it or can we just
> remove it?

It's still needed for some special cases. For example there is no other 
viable way to profile early boot without a VM

I would just drop the hack to unwind, at least for the early boot 
profile use case locking profiling is usually not needed.

-Andi