Re: [Xen-devel] [OSSTEST PATCH 2/2] RFC: enable CONFIG_LIVEPATCH in all relevant arm64 builds

[Julien Grall <julien.grall@xxxxxxx>]

Hi Ian,

On 05/06/2019 12:13, Ian Jackson wrote:
> Right now there is a bug in, at least, Xen 4.11, where it does not
> boot on some of our ARM64 machines but only if CONFIG_LIVEPATCH
> is *dis*abled.
>
> We think this is due to Xen not playing by the proper cache management
> rules in the ARM ARM.  We think that CONFIG_LIVEPATCH is not really
> relevant: it just causes slight changes in timing or memory accesses
> or code layout: so, another roll of the dice.  Ie we think it is
> random coincidence that it works with CONFIG_LIVEPATCH and fails
> without it.
>
> Right now the bug is masked in the osstest flights which test the Xen
> branches because they all are all built with CONFIG_LIVEPATCH enabled.
> The non-Xen branches until now have been built with CONFIG_LIVEPATCH
> disabled.
>
> The overall result is blockage of the CI for, currently,
> qemu-xen-4.11-testing.  This blockage is not a true regression; the
> code was always buggy.  It's just that something has changed and now
> the bug bites.
>
> Switching on CONFIG_LIVEPATCH for the affected tests will, hopefully,
> make this problem "go away" again.  This is clearly a bodge.  But it
> is better than simply force pushing: if we can get it to boot, we will
> be able to run the other tests.
How about disabling the test on rochester?

Cheers,

> I have used standalone-generate-dump-flight-runvars and the effect is
> to turn on enable_livepatch in build-arm64 in precisely
> qemu-upstream-xen-4.11-testing.
>
> AIUI:
>
> An alternative would be to backport 00c96d7742 "xen/arm: mm: Set-up
> page permission for Xen mappings earlier on", to affected Xen stable
> branches.  That would differently permute the cases where things fail.
> In particular on some hardware, including (we think) all hardware we
> actually have, the violation of the architecture cache rules would be
> absolutely irrelevant because of microarchitectural details.  But it
> might cause regressions on other hardware.
>
> CC: Stefano Stabellini <sstabellini@xxxxxxxxxx>
> CC: Jan Beulich <JBeulich@xxxxxxxx>
> CC: Anthony Perard <anthony.perard@xxxxxxxxxx
> CC: Julien Grall <julien.grall@xxxxxxx>
> Signed-off-by: Ian Jackson <ian.jackson@xxxxxxxxxxxxx>
> ---
>   mfi-common | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/mfi-common b/mfi-common
> index b86febda..52a38cb4 100644
> --- a/mfi-common
> +++ b/mfi-common
> @@ -373,7 +373,8 @@ create_build_jobs () {
>           fi
>       fi
>   
> -    if branch_wants_livepatch; then
> +    if branch_wants_livepatch ||
> +          [ $arch = arm64 -a "$xenbranch" = xen-4.11-testing ]; then
>          livepatch_runvars='enable_livepatch=true'
>       fi
>       eval "
--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel