Add QCM2290 / scuba

[konradybcio]

Also applies to QRB2210 / RB1

[MarijnS95]

Here you do have a valid GCC mux value to write... #18 (comment)

[konradybcio]

in short, it's not exactly necessary

[MarijnS95]

But the value is written to a register... should we skip that write entirely or does it simply not blow up on a bogus value?

[lumag]

The value is masked out by the mux masks. 0x9b and 0x220 look correct. So this part looks good.

@andersson could you please merge this?

[MarijnS95]

The values do look correct (in the same ballpark as the others), I'm just complaining reminding about #18 (comment) where it was decided to write a value of 0xfeedbeef & 0x1fff to GCC's mux_reg at 0x62024.

[konradybcio]

hm yeah that seems a bit excessive

[lumag]

I have pushed #24 , it should
make MCCC declarations reflect reality by not requiring dummy &gcc or
NULL entries.

[lumag]

Reviewed-by: Dmitry Baryshkov dmitry.baryshkov@linaro.org

[konradybcio]

As a penalty for not merging this for a long time, more patches! :P

[lumag]

@konradybcio please rebase on top of the current trunk. Thank you.

[konradybcio]

@lumag rebased, compiletested only, added a bugfix

@TravMurav could you please confirm 7180 still works?

[TravMurav]

Sorry, didn't get to testing it earlier, on sc7180 it segfaults trying to call a non-assigned parent measure function afaiu

Reading symbols from ./debugcc...
(gdb) r
Starting program: /home/travler/tmp/code/debugcc/build/debugcc -p sc7180 -a

Program received signal SIGSEGV, Segmentation fault.
0x0000000000000000 in ?? ()
(gdb) bt
#0  0x0000000000000000 in ?? ()
#1  0x0000aaaaaaad36cc in measure_leaf (clk=0xaaaaaab08a98 <sc7180_clocks>, mux=0xaaaaaab08a38 <cpu_cc>) at ../debugcc.c:162
#2  0x0000aaaaaaad3794 in measure (clk=0xaaaaaab08a98 <sc7180_clocks>) at ../debugcc.c:183
#3  0x0000aaaaaaad3ffc in main (argc=4, argv=0xfffffffffa98) at ../debugcc.c:394
(gdb)