target/linux/mvebu/patches-5.10/101-cpufreq-armada-37xx-forbid-cpufreq-for-1.2-GHz-variant.patch

   1 From: =?utf-8?q?Marek_Beh=C3=BAn?= <kabel@kernel.org>
   2 Subject: [PATCH v2] cpufreq: armada-37xx: forbid cpufreq for 1.2 GHz variant
   3 Date: Thu,  1 Jul 2021 00:56:01 +0200
   4
   5 The 1.2 GHz variant of the Armada 3720 SOC is unstable with DVFS: when
   6 the SOC boots, the WTMI firmware sets clocks and AVS values that work
   7 correctly with 1.2 GHz CPU frequency, but random crashes occur once
   8 cpufreq driver starts scaling.
   9
  10 We do not know currently what is the reason:
  11 - it may be that the voltage value for L0 for 1.2 GHz variant provided
  12   by the vendor in the OTP is simply incorrect when scaling is used,
  13 - it may be that some delay is needed somewhere,
  14 - it may be something else.
  15
  16 The most sane solution now seems to be to simply forbid the cpufreq
  17 driver on 1.2 GHz variant.
  18
  19 Signed-off-by: Marek Behún <kabel@kernel.org>
  20 Fixes: 92ce45fb875d ("cpufreq: Add DVFS support for Armada 37xx")
  21 ---
  22  drivers/cpufreq/armada-37xx-cpufreq.c | 6 +++++-
  23  1 file changed, 5 insertions(+), 1 deletion(-)
  24
  25 --- a/drivers/cpufreq/armada-37xx-cpufreq.c
  26 +++ b/drivers/cpufreq/armada-37xx-cpufreq.c
  27 @@ -102,7 +102,11 @@ struct armada_37xx_dvfs {
  28  };
  29
  30  static struct armada_37xx_dvfs armada_37xx_dvfs[] = {
  31 -       {.cpu_freq_max = 1200*1000*1000, .divider = {1, 2, 4, 6} },
  32 +       /*
  33 +        * The cpufreq scaling for 1.2 GHz variant of the SOC is currently
  34 +        * unstable because we do not know how to configure it properly.
  35 +        */
  36 +       /* {.cpu_freq_max = 1200*1000*1000, .divider = {1, 2, 4, 6} }, */
  37         {.cpu_freq_max = 1000*1000*1000, .divider = {1, 2, 4, 5} },
  38         {.cpu_freq_max = 800*1000*1000,  .divider = {1, 2, 3, 4} },
  39         {.cpu_freq_max = 600*1000*1000,  .divider = {2, 4, 5, 6} },