Re: [PATCH v17 03/15] arm64: Introduce prctl() options to control the tagged user addresses ABI
From: Evgenii Stepanov
Date: Mon Jun 17 2019 - 18:04:42 EST
On Mon, Jun 17, 2019 at 10:18 AM Catalin Marinas
> On Mon, Jun 17, 2019 at 09:57:36AM -0700, Evgenii Stepanov wrote:
> > On Mon, Jun 17, 2019 at 6:56 AM Catalin Marinas <catalin.marinas@xxxxxxx> wrote:
> > > On Wed, Jun 12, 2019 at 01:43:20PM +0200, Andrey Konovalov wrote:
> > > > From: Catalin Marinas <catalin.marinas@xxxxxxx>
> > > >
> > > > It is not desirable to relax the ABI to allow tagged user addresses into
> > > > the kernel indiscriminately. This patch introduces a prctl() interface
> > > > for enabling or disabling the tagged ABI with a global sysctl control
> > > > for preventing applications from enabling the relaxed ABI (meant for
> > > > testing user-space prctl() return error checking without reconfiguring
> > > > the kernel). The ABI properties are inherited by threads of the same
> > > > application and fork()'ed children but cleared on execve().
> > > >
> > > > The PR_SET_TAGGED_ADDR_CTRL will be expanded in the future to handle
> > > > MTE-specific settings like imprecise vs precise exceptions.
> > > >
> > > > Signed-off-by: Catalin Marinas <catalin.marinas@xxxxxxx>
> > >
> > > A question for the user-space folk: if an application opts in to this
> > > ABI, would you want the sigcontext.fault_address and/or siginfo.si_addr
> > > to contain the tag? We currently clear it early in the arm64 entry.S but
> > > we could find a way to pass it down if needed.
> > For HWASan this would not be useful because we instrument memory
> > accesses with explicit checks anyway. For MTE, on the other hand, it
> > would be very convenient to know the fault address tag without
> > disassembling the code.
> I could as this differently: does anything break if, once the user
> opts in to TBI, fault_address and/or si_addr have non-zero top byte?
I think it would be fine.
> Alternatively, we could present the original FAR_EL1 register as a
> separate field as we do with ESR_EL1, independently of whether the user
> opted in to TBI or not.