Quoting Taniya Das (2020-07-14 23:54:10) > The display gpll0 branch clock needs to be always left enabled, thus > move the clock ops to _aon branch ops. > > Signed-off-by: Taniya Das <tdas@xxxxxxxxxxxxxx> I wanted to apply this but it needs more commit text. Here's the commit text I wrote. clk: qcom: gcc: Make disp gpll0 branch aon for sc7180/sdm845 The display gpll0 branch clock inside GCC needs to always be enabled. Otherwise the AHB clk (disp_cc_mdss_ahb_clk_src) for the display clk controller (dispcc) will stop clocking while sourcing from gpll0 when this branch inside GCC is turned off during unused clk disabling. We can never turn this branch off because the AHB clk for the display subsystem is needed to read/write any registers inside the display subsystem including clk related ones. This makes this branch a really easy way to turn off AHB access to the display subsystem and cause all sorts of mayhem. Let's just make the clk ops keep the clk enabled forever and ignore any attempts to disable this clk so that dispcc access keep working. Signed-off-by: Taniya Das <tdas@xxxxxxxxxxxxxx> Reported-by: Evan Green <evgreen@xxxxxxxxxxxx> Link: https://lore.kernel.org/r/1594796050-14511-1-git-send-email-tdas@xxxxxxxxxxxxxx Fixes: 17269568f726 ("clk: qcom: Add Global Clock controller (GCC) driver for SC7180") Fixes: 06391eddb60a ("clk: qcom: Add Global Clock controller (GCC) driver for SDM845") > --- > drivers/clk/qcom/gcc-sc7180.c | 2 +- > drivers/clk/qcom/gcc-sdm845.c | 4 ++-- > 2 files changed, 3 insertions(+), 3 deletions(-) > > diff --git a/drivers/clk/qcom/gcc-sc7180.c b/drivers/clk/qcom/gcc-sc7180.c > index ca4383e..538677b 100644 > --- a/drivers/clk/qcom/gcc-sc7180.c > +++ b/drivers/clk/qcom/gcc-sc7180.c > @@ -1061,7 +1061,7 @@ static struct clk_branch gcc_disp_gpll0_clk_src = { > .hw = &gpll0.clkr.hw, > }, > .num_parents = 1, > - .ops = &clk_branch2_ops, > + .ops = &clk_branch2_aon_ops, I'm worried that dispcc may probe before GCC and then this branch is disabled out of boot. Then we'll try to read dispcc registers to determine clk rates and blow up. I hope that doesn't happen and fixing it would need this patch plus the usage of PM clks in the dispcc driver. I suppose if we run into that problem we can use PM clks to make this branch turn on before accessing any dispcc clks.
