Changelog:
r45695 | gpz | 2025-06-10 19:47:44 +0000 (Tue, 10 Jun 2025) | 1 line
added support for "Magic Desk 16" cartridge (Patch by Crystalct). Note that the correct CRTID is 85, not 79 like in the original patch.
------------------------------------------------------------------------
r45694 | rhialto | 2025-06-05 20:46:21 +0000 (Thu, 05 Jun 2025) | 6 lines
PET 8296 $Axxxx video ram treatment.
On the PET 8296, when writing to $Axxxx which writes through to RAM, treat this
as video ram, so we can do beam racing with it like the $8xxx memory.
------------------------------------------------------------------------
r45693 | gpz | 2025-06-02 00:07:01 +0000 (Mon, 02 Jun 2025) | 1 line
added some hints on tde vs autostart
------------------------------------------------------------------------
r45692 | gpz | 2025-06-01 23:15:06 +0000 (Sun, 01 Jun 2025) | 1 line
fix device traps in xplus4 (hopefully), fixes #2143
------------------------------------------------------------------------
r45691 | rhialto | 2025-06-01 13:21:15 +0000 (Sun, 01 Jun 2025) | 19 lines
c1541: fix segfault when not attaching disk image.
When a disk image is attached, vdrive_device_setup() is called to set up the
vdrive struct. This includes memory for the disk error string buffer.
If you do a "c1541 -write foo bar" command (so without attaching a disk image),
c1541 wants to write a DRIVE NOT READY to the error string buffer. This
segfaults because it has not been allocated.
Fix: Call vdrive_device_setup() right after a vdrive is allocated, so it is
always good to use. Also, when freeing the buffers in vdrive_device_shutdown(),
set them to NULL to make sure they are not re-used accidentally.
When vdrive_device_setup() is called again for attaching a disk image, be
paranoid and first call vdrive_device_shutdown() on it.
When detaching an image, call vdrive_device_shutdown() even if no image was
attached. This prevents memory leaks.
Also: replace some 8's by DRIVE_UNIT_MIN.
------------------------------------------------------------------------