[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH 01/10] xen/arm: ffa: Rework firmware discovery





On 22/09/2024 11:00, Julien Grall wrote:
Hi Bertrand,

On 19/09/2024 14:19, Bertrand Marquis wrote:
Rework firmware discovery during probe:
- move prints into the probe
- rename ffa_version to ffa_fw_version as the variable identifies the
   version of the firmware and not the one we support
- add error prints when allocation fail during probe

No functional changes.

Signed-off-by: Bertrand Marquis <bertrand.marquis@xxxxxxx>
---
  xen/arch/arm/tee/ffa.c | 52 +++++++++++++++++++++++++-----------------
  1 file changed, 31 insertions(+), 21 deletions(-)

diff --git a/xen/arch/arm/tee/ffa.c b/xen/arch/arm/tee/ffa.c
index 022089278e1c..7c84aa6aa43d 100644
--- a/xen/arch/arm/tee/ffa.c
+++ b/xen/arch/arm/tee/ffa.c
@@ -71,8 +71,8 @@
  #include "ffa_private.h"
-/* Negotiated FF-A version to use with the SPMC */
-static uint32_t __ro_after_init ffa_version;
+/* Negotiated FF-A version to use with the SPMC, 0 if not there or supported */
+static uint32_t __ro_after_init ffa_fw_version;
  /*
@@ -105,10 +105,7 @@ static bool ffa_get_version(uint32_t *vers)
      arm_smccc_1_2_smc(&arg, &resp);
      if ( resp.a0 == FFA_RET_NOT_SUPPORTED )
-    {
-        gprintk(XENLOG_ERR, "ffa: FFA_VERSION returned not supported\n");
          return false;
-    }
      *vers = resp.a0;
@@ -372,7 +369,7 @@ static int ffa_domain_init(struct domain *d)
      struct ffa_ctx *ctx;
      int ret;
-    if ( !ffa_version )
+    if ( !ffa_fw_version )
          return -ENODEV;
       /*
        * We can't use that last possible domain ID or ffa_get_vm_id() would
@@ -505,6 +502,9 @@ static bool ffa_probe(void)
       */
      BUILD_BUG_ON(PAGE_SIZE != FFA_PAGE_SIZE);
+    printk(XENLOG_INFO "ARM FF-A Mediator version %u.%u\n",
+           FFA_MY_VERSION_MAJOR, FFA_MY_VERSION_MINOR);
 > +>       /*
       * psci_init_smccc() updates this value with what's reported by EL-3
       * or secure world.
@@ -514,25 +514,21 @@ static bool ffa_probe(void)
          printk(XENLOG_ERR
                 "ffa: unsupported SMCCC version %#x (need at least %#x)\n",
                 smccc_ver, ARM_SMCCC_VERSION_1_2);
-        return false;
+        goto err_no_fw;
      }
      if ( !ffa_get_version(&vers) )
-        return false;
+    {
+        gprintk(XENLOG_ERR, "ffa: FFA_VERSION returned not supported\n");

This error message relies on the implementation of ffa_get_version(). It made sense in the previous placement, but here, it seems a little bit odd. So if you want to move the error message, then I think it should be reworded to be more generic.

Maybe: "Cannot retrieve the FFA version".

+        goto err_no_fw;
+    }
      if ( vers < FFA_MIN_SPMC_VERSION || vers > FFA_MY_VERSION )
      {
          printk(XENLOG_ERR "ffa: Incompatible version %#x found\n", vers);
-        return false;
+        goto err_no_fw;
      }
-    major_vers = (vers >> FFA_VERSION_MAJOR_SHIFT) & FFA_VERSION_MAJOR_MASK;
-    minor_vers = vers & FFA_VERSION_MINOR_MASK;
-    printk(XENLOG_INFO "ARM FF-A Mediator version %u.%u\n",
-           FFA_MY_VERSION_MAJOR, FFA_MY_VERSION_MINOR);

I kind of understand why we are moving the Medatior version early but...

-    printk(XENLOG_INFO "ARM FF-A Firmware version %u.%u\n",
-           major_vers, minor_vers);

... I am not sure why we would move this print later. Wouldn't this be useful to know if there is a missing feature?

Ah so in patch #2 you re-ordered the code again and it looks the same as today. It would be better if we can avoid moving the code completely.

Cheers,

--
Julien Grall




 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.