Re: Allwinner D1

[Rui-Xiang Guo <rxg%lavabit.com@localhost>]

On Wed, Aug 21, 2024 at 02:48:56PM +0900, Rin Okuyama wrote:
> On 2024/08/18 21:35, Rui-Xiang Guo wrote:
> > > > Now I have trouble with attaching sunximmc which is from arm/sunxi/sunxi_mmc.c:
> > 
> > Updated tail:
> > [   1.0000030] sunximmc0: host reset
> > [   1.0000030] sunximmc0: width = 1
> > [   1.0000030] sdmmc0 at sunximmc0
> > [   1.0000030] sdmmc1 at sunximmc1
> > [   1.0000030] sunximmc0: update clock
> > [   1.0000030] sunximmc0: update clock
> > [   1.0000030] sunximmc0: update clock
> > [   1.0000030] WARNING: system needs entropy for security; see entropy(7)
> > 
> > It is stuck in kpause(), any ideas?
> 
> Full backtrace should be helpful.

I am attempting to add sunxitimer via riscv_timer_register(), and then get
the output:
[...]
[   1.0000030] sunxitimer0: interrupting on plic0 irq 75
[   1.0023310] sunximmc0: host reset
[   1.0023310] sunximmc0: width = 1
[   1.0027578] sdmmc0 at sunximmc0
[   1.9569563] sdmmc1 at sunximmc1
[   1.9600358] sunximmc0: update clock
[   2.9153101] sunximmc0: update clock
[   2.9153101] sunximmc0: update clock
[   4.8278825] WARNING: system needs entropy for security; see entropy(7)
[   8.6557525] panic: kernel diagnostic assertion "p != NULL" failed: file "/usr/src/sys/kern/kern_clock.c", line 446
[   8.6557525] cpu0: Begin traceback...
[   8.6557525] have_addr: true
[   8.6557525] addr: ffffffc04e1ebb60
[   8.6557525] count: 65535
[   8.6557525] modif:
[   8.6557525] trace fp ffffffc04e1ebb60
[   8.6557525] fp ffffffc04e1ebba0 vpanic() at ffffffc000329128 netbsd:vpanic+0x110
[   8.6557525] fp ffffffc04e1ebbc0 kern_assert() at ffffffc000446e92 netbsd:kern_assert+0x38
[   8.6557525] fp ffffffc04e1ebc40 statclock() at ffffffc0002c3a12 netbsd:statclock+0x9c
[   8.6557525] fp ffffffc04e1ebc80 hardclock() at ffffffc0002c3c5a netbsd:hardclock+0x98
[   8.6557525] fp ffffffc04e1ebcb0 sunxi_timer_intr() at ffffffc000007f5e netbsd:sunxi_timer_intr+0x5a
[   8.6557525] fp ffffffc04e1ebcf0 plic_intr() at ffffffc0000185cc netbsd:plic_intr+0x46
[   8.6557525] fp ffffffc04e1ebd60 intc_intr_handler() at ffffffc000019688 netbsd:intc_intr_handler+0xa8
[   8.6557525] tf ffffffc04e1ebd60 exception_kernexit() at ffffffc00000c1cc netbsd:cpu_exception_handler+0x98


> In case you are not familiar with ddb(4):
> 
> (1) Enter DDB from console, by sending break signal, or arbitrary
> key sequence specified by kernel config:
> 
> ----
> options        CNMAGIC="\"+++++\""
> ----
> 
> (2) `ps` command lists kernel/user threads.
> 
> (3) `bt/a (address of struct lwp *, shown in ps)` shows backtrace
> for the target thread.
> 
> BTW:
> Sorry if I'm going too fast, but have you ever tried USB? Both
> OTG and EHCI advertise these are compatible to our drivers:
> 
> https://nxr.netbsd.org/xref/src/sys/arch/arm/sunxi/sunxi_musb.c#67
> https://nxr.netbsd.org/xref/src/sys/dev/fdt/ehci_fdt.c#56

The USB part has not been implemented yet in sun20i_d1_ccu.c,
I will try to add it later.

-rxg