m65 PRG injection load address gets coerced unexpectedly

[dansanderson]

Test Environment (required)

• Tools Platform: MacOS
• Tools Version: 20221008.12-refact-093eae9
• MEGA65 Platform: MEGA65
• ROM Release: 920377
• Core Commit: 2f545d1

Describe the bug When I run m65 myfile.prg, it reads the load address, then types on the MEGA65 REM M65 INJECTING AT {load_addr}. It then tests the load address to see if it's a SID file, and if not, it overrides the load address based on command-line switches. If my PRG file starts at $3000, m65 announces that it is loading at $3000, but actually loads at $2001.

There is a command line argument, -1, that stops this override and honors the load address. However, this argument is not documented under -h help (at least in the version I'm using).

Fix 1: Report the injection address accurately Regardless of other changes, m65 should not misreport the injection address. REM M65 INJECTING AT {load_addr} should state the actual injection address whether it is derived, overridden, or specified on the command line.

Fix 2: Document the -1 flag If the -1 flag is to be the way the user tells m65 to honor the load address of the PRG, it should be documented in the -h help text.

Fix 3: Make -1 the default behavior I can't think of a reason why m65 should override the load address unless I specifically tell it to with the -@ flag. As written, m65 assumes that any PRG that doesn't have one of two SID load addresses is either a C64 BASIC PRG if the -4 argument is provided or a C65 BASIC PRG if the -1 argument is not provided. An actual BASIC PRG should have the correct load address in its header. I believe this block should be removed.

If there's a use case I'm missing and -@ doesn't meet that case, I would expect "force a BASIC load address" to be a separate flag that defaults to off and decides which address to use from the presence of the -4 flag.

[lydon42]

I agree that -1 should be in the help and that the printed text should match the address.

Everything else changes the behavior of the switches, which we are trying to avoid.

[lydon42]

EDIT: the following is wrong, it works concurrently.

One additional problem: You can use a filename and -@ at the same time. But -@ will use overwrite the filename specified for the normal load, so you can't really use both. This is an artifact from the original confusing flow of growing m65 tool options. I think it would be nice if both work concurrently... so this is also an enhancement.