On 31/01/18 17:39, Stefano Stabellini wrote:
On Wed, 31 Jan 2018, Julien Grall wrote:
Hi,
On 31/01/18 16:53, Julien Grall wrote:
+GLOBAL(hyp_traps_vector_bp_inv)
+/*
+ * We encode the exception entry in the bottom 3 bits of
+ * SP, and we have to guarantee to be 8 byt
On Wed, 31 Jan 2018, Julien Grall wrote:
> Hi,
>
> On 31/01/18 16:53, Julien Grall wrote:
> > +GLOBAL(hyp_traps_vector_bp_inv)
> > +/*
> > + * We encode the exception entry in the bottom 3 bits of
> > + * SP, and we have to guarantee to be 8 bytes aligned.
> > + */
Hi,
On 31/01/18 16:53, Julien Grall wrote:
+GLOBAL(hyp_traps_vector_bp_inv)
+/*
+ * We encode the exception entry in the bottom 3 bits of
+ * SP, and we have to guarantee to be 8 bytes aligned.
+ */
+add sp, sp, #1 /* Reset7 */
From: Julien Grall
In order to avoid aliasing attackes agains the branch predictor, let's
invalidate the BTB on guest exist. This is made complicated by the fact
that we cannot take a branch invalidating the BTB.
This is based on the first version posrted by Marc Zyngier on Linux-arm
mailing lis
