Skip to content

dbdma: Interpret command lists until STOP - #220

Merged
dingusdev merged 2 commits into
dingusdev:masterfrom
mihaip:upstream-classic-dma
Aug 22, 2026
Merged

dbdma: Interpret command lists until STOP#220
dingusdev merged 2 commits into
dingusdev:masterfrom
mihaip:upstream-classic-dma

Conversation

@mihaip

@mihaip mihaip commented Aug 22, 2026

Copy link
Copy Markdown
Contributor

With #214 we started to advertise the IDE DMA support from #192 to guests, which appeared to break booting Mac OS 8.x (at least on the Beige G3) - the startup disk was not even detected. It looks like this is because the DMA is enabled, it gets translated to this sequence of descriptors:
1.INPUT_MORE
2.NOP (with the i interrupt bit set)
3. STOP (which clears the active bit set)

The emulated transfer completion command executed only one following descriptor and stopped at the NOP. This generated the interrupt and left the channel active, and thus in an inconsistent state. The Mac OS X driver is more tolerant of this, which is why it wasn't an issue there.

The DBDMA specification says "The target fetches command entries and performs the command-entry-specified data-transfer operation, processing command entries until a STOP entry is reached", so a loop is more correct (and we were already doing it for start and resume`, and empirically it allows 8.1 to boot again.

As part of fixing this I also noticed that we had some remnants of the old DMA implementation, to remove them.

mihaip added 2 commits August 21, 2026 22:31
With dingusdev#214 we started to advertise the IDE DMA support from dingusdev#192 to
guests. That appeared to break booting Mac OS 8.x, at least on the Beige
G3: the startup disk was not even detected.

Mac OS 8.1 sets up an IDE DMA read as INPUT_MORE, then a NOP with its i
field set to always, then STOP, which clears the channel's ACTIVE bit.
The emulated transfer completion path interpreted only one following
descriptor, so it stopped at NOP. NOP raised the interrupt, but STOP was
never reached and the channel stayed active. The Mac OS X driver is more
tolerant of this, which is why it was not an issue there.

The DBDMA Specification, section 1.4, says: "The target fetches command
entries and performs the command-entry-specified data-transfer operation,
processing command entries until a STOP entry is reached". Keep
interpreting ready commands until a transfer blocks or the channel stops,
as start() and resume() already did. This is closer to the hardware
behavior, and empirically lets Mac OS 8.1 boot again.
b63b658 removed the old ESCC DMA implementation, taking the last users
of end_pull_data(), end_push_data(), and get_push_data_remaining() with
it. 9fb34ef later removed the last get_pull_data_remaining() call from
the sound server.

Nothing has called these hooks since. Remove them from DmaOutChannel,
DmaInChannel, and DMAChannel instead of carrying dead completion paths
alongside the DMA interfaces that are still in use.
@mihaip mihaip changed the title Upstream classic dma dbdma: Interpret command lists until STOP Aug 22, 2026
@dingusdev
dingusdev merged commit 4d3cd65 into dingusdev:master Aug 22, 2026
7 checks passed
@mihaip
mihaip deleted the upstream-classic-dma branch August 23, 2026 23:40
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants