doc: add reverse engineering notes for trs80-4p

Lines should be terminated only with a CR on this system, not CRLF.

doc/trs80-4p.md

@@ -210,3 +210,34 @@ always successful.
     66  BANK    B=func C=bank       BZ   Memory bank use
     67  BREAK   HL=vector           HL   Set Break vector
     68  SOUND   B=func              -    Sound generation
+
+## Personal reverse engineering
+
+This section below contains notes about my personal reverse engineering efforts.
+I'm not an expert in this, and also, I might not be aware of existing, better
+documentation making this information useless.
+
+### Bootable disk
+
+I'm wondering what makes a disk bootable to the TRS-80 and how it boots it.
+When I read the raw contents of the first sector of the first cylinder of the
+TRS-DOS disk, I see that, except for the 3 first bytes (`00fe14`), the rest of
+the contents is exactly the same as what is at memory offset `0x0203`, which
+seems to indicates that the bootloader simply loads that contents to memory,
+leaving the first 3 bytes of RAM to either random contents or some predefined
+value (I have `f8f800`).
+
+A non-bootable disk starts with `00fe14`, but we can see the message "Cannot
+boot, DA TA DISK!" at offset `0x2a`.
+
+I'm not sure what `00fe14` can mean. Disassembled, it's
+`nop \ rst 0x28 \ ld b, c`. It makes sense that booting would start with a
+service call with parameters set by the bootloader (so we don't know what that
+service call actually is), but I'm not sure it's what happens.
+
+I don't see any reference to the `0x2a` offset in the data from the first
+sector, but anyways, booting with the non-bootable disk doesn't actually prints
+the aformentioned message, so it might be a wild goose chase.
+
+In any case, making a disk bootable isn't a concern as long as Collapse OS uses
+the TRS-DOS drivers.

kernel/stdio.asm

@@ -76,6 +76,9 @@ printnstr:
 	pop	bc
 	ret
 
+; Prints a line terminator. This routine is a bit of a misnomer because it's
+; designed to be overridable to, for example, printlf, but we'll live with it
+; for now...
 printcrlf:
 	push	af
 	ld	a, CR

recipes/trs80/README.md

@@ -151,6 +151,10 @@ A memory range dumped this way will be re-loaded at the same offset through
 using the `RUN` command. Therefore, you can avoid all this work above in later
 sessions by simply typing `recv` in the DOS prompt.
 
+Note that you might want to turn `debug` off for these commands to run. I'm not
+sure why, but when the debugger is on, launching the command triggers the
+debugger.
+
 ## Sending binary through the RS-232 port
 
 Once you're finished punching your program in memory, you can run it with
@@ -186,6 +190,12 @@ If there was no error during `pingpong`, the content should be exact.
 Nevertheless, I recommend that you manually validate a few bytes using TRSDOS
 debugger before carrying on.
 
+*debugging tip*: Sometimes, the communication channel can be a bit stubborn and
+always fail, as if some leftover data was consistently blocking the channel. It
+would cause a data mismatch at the very beginning of the process, all the time.
+What I do in these cases is start a `COMM *cl` session on one side and a screen
+session on the other, type a few characters, and try `pingpong` again.
+
 ## Running Collapse OS
 
 If everything went well, you can run Collapse OS with `g3000<space>`. You'll

recipes/trs80/glue.asm

@@ -18,6 +18,9 @@
 .equ	STDIO_PUTC	trs80PutC
 .inc "stdio.asm"
 
+; The TRS-80 generates a double line feed if we give it both CR and LF.
+.equ	printcrlf	printcr
+
 ; *** BASIC ***
 
 ; RAM space used in different routines for short term processing.
@@ -44,4 +47,11 @@ init:
 	call	basInit
 	jp	basStart
 
+printcr:
+	push	af
+	ld	a, CR
+	call	STDIO_PUTC
+	pop	af
+	ret
+
 RAMSTART: