It was <2024-06-21 pią 14:40>, when Sam Protsenko wrote:

On Fri, Jun 21, 2024 at 2:00 PM Lukasz Stelmach [off-list ref] wrote:

quoted

It was <2024-06-20 czw 18:13>, when Sam Protsenko wrote:

quoted

On some Exynos chips like Exynos850 the access to Security Sub System
(SSS) registers is protected with TrustZone, and therefore only possible
from EL3 monitor software. The Linux kernel is running in EL1, so the
only way for the driver to obtain TRNG data is via SMC calls to EL3
monitor. Implement such SMC operation and use it when EXYNOS_SMC flag is
set in the corresponding chip driver data.

Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Krzysztof Kozlowski <redacted>
---
Changes in v3:
  - Added appropriate error messages for the case when init SMC command fails

Changes in v2:
  - Used the "reversed Christmas tree" style in the variable declaration
    block in exynos_trng_do_read_smc()
  - Renamed .quirks to .flags in the driver structure
  - Added Krzysztof's R-b tag

 drivers/char/hw_random/exynos-trng.c | 140 +++++++++++++++++++++++++--
 1 file changed, 130 insertions(+), 10 deletions(-)

diff --git a/drivers/char/hw_random/exynos-trng.c b/drivers/char/hw_random/exynos-trng.c
index 6ef2ee6c9804..9fa30583cc86 100644
--- a/drivers/char/hw_random/exynos-trng.c
+++ b/drivers/char/hw_random/exynos-trng.c

[...]

quoted

@@ -103,6 +163,24 @@ static int exynos_trng_init(struct hwrng *rng)
      return 0;
 }

+static int exynos_trng_init_smc(struct hwrng *rng)
+{
+     struct exynos_trng_dev *trng = (struct exynos_trng_dev *)rng->priv;
+     struct arm_smccc_res res;
+     int ret = 0;
+
+     arm_smccc_smc(SMC_CMD_RANDOM, HWRNG_INIT, 0, 0, 0, 0, 0, 0, &res);
+     if (res.a0 != HWRNG_RET_OK) {
+             dev_err(trng->dev, "SMC command for TRNG init failed (%d)\n",
+                     (int)res.a0);
+             ret = -EIO;
+     }
+     if ((int)res.a0 == -1)
+             dev_info(trng->dev, "Make sure LDFW is loaded by your BL\n");

This is good, thank you for adding it. It can be even better though, if
you don't skimp on message length (-; I mean, I know what BL is, I can
fingure what LDFW is because you have explained to me and I can see the
source code, but somewone who sees it for the first time will be only
slightly less surprised than with v2 error message only. Come on, you
can make this message twice as long and it will still fit in 80 characters (-;

Guess my OCD got in the way and I just didn't want to break the line
:) But yeah, LDFW = Loadable Firmware, and BL = bootloader. There is
an "ldfw" partition on eMMC, and I noticed Samsung usually uses LDFW
term, so I figured it was not a big deal to throw that abbreviation at
the user. But I totally agree on BL part, it might be confusing. I
don't have any strong opinion on this one. If you are going to apply
v3, can I kindly ask you to change that message the way you want it to
be?

I guess Olivia or Herbert will be applying it. Let me try… How about:

"Check if your bootloader loads the firmware (SMC) part of the driver."

quoted

Don't change it if v3 is the last. If not, please, make it more verbose.

quoted

+
+     return ret;
+}
+

[...]


Kind regards,
--
Łukasz Stelmach
Samsung R&D Institute Poland
Samsung Electronics

-- 
Łukasz Stelmach
Samsung R&D Institute Poland
Samsung Electronics