Re: [PATCH] arm64: mm: support instruction SETEND

From: Will Deacon
Date: Wed Jan 07 2015 - 06:12:02 EST


On Wed, Jan 07, 2015 at 10:58:24AM +0000, Leo Yan wrote:
> On Wed, Jan 07, 2015 at 10:10:34AM +0000, Suzuki K. Poulose wrote:
> > On 07/01/15 05:52, Leo Yan wrote:
> > >Currently kernel has set the bit SCTLR_EL1.SED, so the SETEND
> > >instruction will be treated as UNALLOCATED; this error can be
> > >reproduced when ARMv8 cpu runs with EL1/aarch64 and EL0/aarch32
> > >mode, finally kernel will trap the exception if the userspace
> > >libs use SETEND instruction.
> > >
> > >So this patch clears bit SCTLR_EL1.SED to support SETEND instruction.
> > >
> > The best way to do this, is via the instruction emulation framework
> > added by Punit, which handles the armv8 deprecated/obsoleted
> > instructions. This is now queued for 3.19.
> > I have a patchset which adds the 'SETEND' emulation support to the
> > framework. This will enable better handling of the feature,
> > including finding out the users of the deprecated instruction (when
> > we switch to the emulation mode).
> >
>
> i'm a little confuse for this point:
>
> if the deprecated instructions cannot be supported by CPU, then only
> can use emulation; on the other hand, if CPU can natively support the
> deprecated instruction, why we cannot directly enable this h/w feature?
> if use the emulation mode, suppose here will have performance penalty.
>
> how about u think for this? :-)

A *huge* advantage of emulation is that we can print a diagnostic to dmesg
warning the user that they are making use of a CPU feature that is likely to
disappear from future revisions of the hardware. We've not been great at
doing this in the past, which led to all the fun around SWP emulation that
you can find in the list archives.

Furthermore, I think the emulation framework does allow the hardware support
to be enabled, it just doesn't make that the default behaviour.

In other words; use the emulation to find out where SETEND is being used
and fix those applications wherever you can. In the cases where you're
stuck with a legacy binary, you can enable CPU support if it is available
but that's not a longterm solution.

Will
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/