Hi Roman,
On 06/01/2021 11:26, Roman Skakun wrote:
This patch added additional sanity and increases an understanding for
getting proper value from the first argument for SMC call on aarch64
according to SMCC Convention.
I would suggest the following commit message:
"xen/arm: optee: The function identifier is always 32-bit
Per the SMCCC specification (see section 3.1 in ARM DEN 0028D), the
function identifier is only stored in the least significant 32-bits. The
most significant 32-bits should be ignored.
The function optee_handle_call() is now updated to ignore the most
significant 32-bits.
"
Note that I used the version D rather than B because the latter has
buggy wording (it seems to suggest that the least significants bits
should be ignored).
Furthermore, I checked vsmc.c (the layer that dispatch the SMC) and it
was already handled properly thanks to commit 7f4217cc6057 "xen/arm:
vsmc: The function identifier is always 32-bit".
[0] ARM DEN0028B, page 12
Signed-off-by: Roman Skakun <roman_ska...@epam.com>
---
xen/arch/arm/tee/optee.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/xen/arch/arm/tee/optee.c b/xen/arch/arm/tee/optee.c
index ee85359742..87060b52b8 100644
--- a/xen/arch/arm/tee/optee.c
+++ b/xen/arch/arm/tee/optee.c
@@ -1643,7 +1643,8 @@ static bool optee_handle_call(struct cpu_user_regs *regs)
if ( !ctx )
return false;
- switch ( get_user_reg(regs, 0) )
+ /* Only least 32 bits are significant (see ARM DEN 0028B, page 12) */
I would suggest:
/* The function identifier is always stored in the least significant
32-bit (see section ARM DEN 0028D) */
I can update it while committing, if both Volodymyr and you are happy
with changes.
Cheers,
+ switch ( (uint32_t)get_user_reg(regs, 0) )
{
case OPTEE_SMC_CALLS_COUNT:
set_user_reg(regs, 0, OPTEE_MEDIATOR_SMC_COUNT);
--
Julien Grall
