The current version can be found at:
https://bitbucket.org/superfury/unipcemu/src/ … ide.c?at=master

The driver (vide-cdd.sys) doesn't seem to read the result from the buffer for some reason after transferring the complete command?

After the latest CPU OUTSB/W fix, now the correct packet is loaded into the ATAPI4 CD-ROM drive. Then it starts to infinitely read the Status Register(IRQ is asserted directly when the packet is fully received, commandstatus becoming 1(Indicating it's to read the packet result from the device)). The LBA registers in total contain 0xA0240001(Sector number 0x01, Size 0x24).

Strangely I eventually notice that the status returns to 0(Which usually is the case because of an invalid command or device reset)?
https://bitbucket.org/superfury/unipcemu/src/ … ide.c?at=master

After the latest changes in the 80286 emulation, the OAKCDROM.SYS doesn't seem to crash just like that anymore. Now it seems to be doing something strange: it's outputting some byte to the ATA data port, then reading the status register, expecting it to not set bits 6&4(!=0x50)? Why would it do this? Is some kind of ready/not ready behaviour expected and required for CD-ROM drivers?

The controller itself is in the default command status(status 0, which means ready for command), ready to accept a command. The status register contains 0x40.

Anyone can see why the cd-rom driver fails to continue after sending the inquiry command with parameters? Why isn't it reading the result?

https://bitbucket.org/superfury/unipcemu/src/ … ide.c?at=master

It's clearing the IDE busy flag, since it's not doing anything while transferring packet data(the result data that's in the buffers, or transferring packet data(like Mode data tables)). Also no sector count register is set(it's in polling mode always). Is that a problem with CD-ROM drivers?

Current source code: https://bitbucket.org/superfury/unipcemu/src/ … ide.c?at=master