[integrated,sw] Switch logging over to dbg_print

[sboeuf]

Log traces with dbg_print instead of rom_print.
dbg_print comes from OpenTitan master branch and rom_print has been
deprecated.

[rswarbrick]

This looks reasonable to me, but I think it would probably make sense for a SW person to take a look. I'm adding Chris Frantz as a reviewer, because I think he wrote dbg_print.c in the first place.

[cfrantz]

I have no objection to you making this change, but please consider the following caveats:

1. This printf implementation is not hardened. With the exception of %s, it will never follow a pointer, but there is more that can be done to harden the implementation. There will be a hardening exercise done on this implementation at some point in the future.
2. The Earlgrey ROM is careful to exclude any of the functions from the dbg_print module. There is an exception for FPGA bitstreams (ie: printing ROM:%x), but because of (1), Earlgrey does not want this function in the ROM.
3. dbg_printf is around 500 bytes. If you are size constrained and you don't need a general printing function, you're better off crafting your output functions from uart_write and/or uart_putchar.

[sameo]

I have no objection to you making this change, but please consider the following caveats:

1. This printf implementation is not hardened.  With the exception of `%s`, it will never follow a pointer, but there is more that can be done to harden the implementation.

Right, and I believe this is not a regression compared to the rom_print version.

There will be a hardening exercise done on this implementation at some point in the future.

Good, we will port that part as well.

3. `dbg_printf` is around 500 bytes.  If you are size constrained and you don't need a general printing function, you're better off crafting your output functions from `uart_write` and/or `uart_putchar`.

Thanks for the heads up. We are not that constrained. Yet...

[andreaskurth]

Approving based on comments above. Thx!