Change in coreboot[master]: console/flashsconsole: Adding a spi flash console for debugging

[Julius Werner (Code Review)]

Julius Werner has posted comments on this change. ( https://review.coreboot.org/19849 )

Change subject: console/flashsconsole: Adding a spi flash console for debugging
......................................................................

Patch Set 2:

(9 comments)

https://review.coreboot.org/#/c/19849/2/src/console/Kconfig
File src/console/Kconfig:

PS2, Line 242: region
It's called "file" in CBFS (and I think "section" is more common for FMAP).

Line 258: 	  region in the CBFS.
Why define two mechanisms? The FMAP seems like a more appropriate place to put this, and all boards use FMAPs these days, so why not just standardize on that? (You can still have a Kconfig for the size to be used in the default FMAP.)

Line 260: if CONSOLE_SPI_FLASH
Just use depends on instead of this block.

https://review.coreboot.org/#/c/19849/2/src/drivers/spi/flashconsole.c
File src/drivers/spi/flashconsole.c:

Line 41: 	if (cbfs_boot_locate(&file, "console", NULL) == 0) {
nit: probably looks better with

if (thing_that_can_have_errors()) {
   printk(...error...);
   return;
 }

It might also help you not go over the 80 char limit, which you do a lot in this function (or is that my Gerrit being weird?).

Line 55: 		boot_device_init();
Already part of prepare_program_locate

Line 60: 		 * the end of a previous log write.
You're banking on the fact that console output can never have legitimate 0xFFs. I think it would be safer to read back from the end of the buffer and find the first byte that's not 0xFF.

Line 121: 		for (i = 0; i < LINE_BUFFER_SIZE - 1; i++) {
Why not just store the current index in a global as well, rather than searching for it all the time?

Line 129: 		if (len >= LINE_BUFFER_SIZE - 1 ||
Why not tie this to tx_flush instead?

https://review.coreboot.org/#/c/19849/2/src/include/console/flash.h
File src/include/console/flash.h:

Line 25: #define __CONSOLE_FLASH_ENABLE__	IS_ENABLED(CONFIG_CONSOLE_SPI_FLASH)
Why are you redefining this? Makes it harder to see that it's governed by a Kconfig.