Re: [RFCv2 00/10] Linear Address Masking enabling

From: Thomas Gleixner <tglx@linutronix.de>
To: "H.J. Lu" <hjl.tools@gmail.com>, Dave Hansen <dave.hansen@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
	"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	Andy Lutomirski <luto@kernel.org>,
	the arch/x86 maintainers <x86@kernel.org>,
	Alexander Potapenko <glider@google.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	Andi Kleen <ak@linux.intel.com>,
	Rick Edgecombe <rick.p.edgecombe@intel.com>,
	Linux-MM <linux-mm@kvack.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [RFCv2 00/10] Linear Address Masking enabling
Date: Fri, 13 May 2022 01:35:16 +0200	[thread overview]
Message-ID: <87o802tjd7.ffs@tglx> (raw)
In-Reply-To: <CAMe9rOqb6ZnAZYe4uAWDt-vmhhP=z_+uZwi5fBURqyUWxCX9Cg@mail.gmail.com>

On Thu, May 12 2022 at 15:10, H. J. Lu wrote:
> On Thu, May 12, 2022 at 2:51 PM Dave Hansen <dave.hansen@intel.com> wrote:
>> On 5/12/22 12:39, Thomas Gleixner wrote:
>> >> It's OK for a debugging build that runs on one kind of hardware.  But,
>> >> if we want LAM-using binaries to be portable, we have to do something
>> >> different.
>> >>
>> >> One of the stated reasons for adding LAM hardware is that folks want to
>> >> use sanitizers outside of debugging environments.  To me, that means
>> >> that LAM is something that the same binary might run with or without.
>> > On/off yes, but is there an actual use case where such a mechanism would
>> > at start time dynamically chose the number of bits?
>>
>> I'd love to hear from folks doing the userspace side of this.  Will
>> userspace be saying: "Give me all the bits you can!".  Or, will it
>> really just be looking for 6 bits only, and it doesn't care whether it
>> gets 6 or 15, it will use only 6?
>>
>> Do the sanitizers have more overhead with more bits?  Or *less* overhead
>> because they can store more metadata in the pointers?
>>
>> Will anyone care about the difference about potentially missing 1/64
>> issues with U57 versus 1/32768 with U48?
>
> The only LAM usage I know so far is LAM_U57 in HWASAN.

That's at least a halfways useful answer.

> An application can ask for LAM_U48 or LAM_U57. But the decision should
> be made by application.

It can ask for whatever, but the decision whether it's granted is made
by the kernel for obvious reasons.

> When an application asks for LAM_U57, I expect it will store tags in
> upper 6 bits, even if the kernel enables LAM_U48.

The kernel does not enable LAM_U48 when the application only wants to
have LAM_U57, because that would restrict the address space of the
application to 47 bits on 5-level capable system for no reason.

So what are you trying to tell me?

Thanks,

        tglx