[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v2 2/3] grackle: use qdev gpios for PCI IRQs
From: |
Mark Cave-Ayland |
Subject: |
Re: [PATCH v2 2/3] grackle: use qdev gpios for PCI IRQs |
Date: |
Sat, 17 Oct 2020 14:03:55 +0100 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.3.1 |
On 16/10/2020 01:26, BALATON Zoltan via wrote:
As said in previous message the i2c and SPD patches are not quite ready yet so I've
omitted those from this series, I may rework them later once this part is merged and
can rebase the rest on top of that. We would also need your screamer patches to get
the Mac ROM working, what is still missing for those?
The 2 main reasons for not merging the screamer patches so far are:
1) Hangs in MacOS 9.0 and 9.1 on startup
Probably related to DBDMA interrupts, but I haven't had time to dig into this in much
detail.
2) Reduced OS X emulation speed
When OS X detects the sound hardware it enables its internal sound engine which does
2 things: firstly it constantly runs DBDMA requests which execute in the bottom-half
even if no sound is being generated, so you end up reducing the raw emulation speed
and secondly the OS X sound engine is floating point based so you end up running a
lot more background floating point arithmetic in the OS.
I'm open to further ideas as to how this can be improved. The DBDMA overhead could be
reduced by running DBDMA in the iothread if that is possible but that would be a fair
bit of work.
ATB,
Mark.
Re: [PATCH v2 2/3] grackle: use qdev gpios for PCI IRQs, David Gibson, 2020/10/15
[PATCH v2 3/3] uninorth: use qdev gpios for PCI IRQs, Mark Cave-Ayland, 2020/10/13
Re: [PATCH v2 3/3] uninorth: use qdev gpios for PCI IRQs, Philippe Mathieu-Daudé, 2020/10/13