This patchset significantly improves OpenBIOS's ability to boot OS 9, adding
ESCC legacy ports, reworking ADB interaction with QEMU and fixing IDE
controller interaction with the OS.
Note that the patch to interpreter.fs is a temporary measure pending a fix
that Mark is writing, and that RTAS must be disabled for the OS to boot.
Signed-off-by: Cormac O'Brien <cormac(a)c-obrien.org>
---
Changes from v1:
* Initialize ESCC legacy ports
* Don't enable non-implemented DBDMA interrupts
* Update some commit messages and comments
Cormac O'Brien (5):
interpreter.fs: allow evaluate to split words on CR as well as LF
ppc: fix up power-mgt properties
ppc: add ESCC legacy support
ppc: fix CUDA ADB packet header format
ppc: fix up IDE device tree and control transfer
arch/ppc/qemu/methods.c | 2 +
drivers/cuda.c | 17 ++++++--
drivers/escc.c | 88 ++++++++++++++++++++++++++++++++++--------
drivers/escc.h | 2 +
drivers/ide.c | 34 ++++++++++++++--
drivers/pci.c | 6 +++
forth/bootstrap/interpreter.fs | 2 +-
7 files changed, 126 insertions(+), 25 deletions(-)
--
2.5.0
On Aug 20, 2015, at 3:20 PM, Cormac O'Brien wrote:
> On Thu, Aug 20, 2015 at 11:17 AM, Programmingkid
> <programmingkidx(a)gmail.com> wrote:
>>
>> On Aug 19, 2015, at 5:48 PM, Mark Cave-Ayland wrote:
>>
>>> On 19/08/15 22:45, Programmingkid wrote:
>>>
>>>> On Aug 19, 2015, at 5:39 PM, Mark Cave-Ayland wrote:
>>>>
>>>>> On 19/08/15 22:31, Programmingkid wrote:
>>>>>
>>>>>> Would you know how to disable the RTAS? I want to test out these patches but no directions were given on how to disable the node.
>>>>>
>>>>> Looks like you just missed my latest email - there should be no change
>>>>> required for the RTAS settings.
>>>>
>>>> Ok. If you have tested Cormac's patches, do you see Mac OS 9 booting? I followed all the directions from his testing page and couldn't see anything that belongs to Mac OS 9. Here is the url: http://c-obrien.org/qemu-os9/testing/
>>>>
>>>> I did my test on Mac OS X, so that could be the problem.
>>>
>>> Yes I did have it working with the github macos9 branch and the Cormac's
>>> github upstream branch earlier in the week. Note that sometimes you can
>>> get a hang before the grey screen appears...
>>>
>>> If you have a specific problem, reply to the thread on the mailing list
>>> and supply a link to a screenshot so other people can test.
>>>
>>
>>
>> I have followed the directions twice from the testing page and have ended up with just this error message:
>> MacOS: unable to find a usable NVRAM partititon - using offset 0x1400.
>>
>> I am building on Mac OS X, so that could be the problem.
>
> That message is itself not an error, just informational -- OS 9 will
> reserve some memory to function as the missing NVRAM. It takes a
> varying amount of time to continue from there but it should be less
> than 10 seconds.
I let QEMU run for several minutes after the message and never saw anything else. I guess something is different with the Mac OS X build process that prevents Mac OS 9 from working.
On Aug 19, 2015, at 5:48 PM, Mark Cave-Ayland wrote:
> On 19/08/15 22:45, Programmingkid wrote:
>
>> On Aug 19, 2015, at 5:39 PM, Mark Cave-Ayland wrote:
>>
>>> On 19/08/15 22:31, Programmingkid wrote:
>>>
>>>> Would you know how to disable the RTAS? I want to test out these patches but no directions were given on how to disable the node.
>>>
>>> Looks like you just missed my latest email - there should be no change
>>> required for the RTAS settings.
>>
>> Ok. If you have tested Cormac's patches, do you see Mac OS 9 booting? I followed all the directions from his testing page and couldn't see anything that belongs to Mac OS 9. Here is the url: http://c-obrien.org/qemu-os9/testing/
>>
>> I did my test on Mac OS X, so that could be the problem.
>
> Yes I did have it working with the github macos9 branch and the Cormac's
> github upstream branch earlier in the week. Note that sometimes you can
> get a hang before the grey screen appears...
>
> If you have a specific problem, reply to the thread on the mailing list
> and supply a link to a screenshot so other people can test.
>
I have followed the directions twice from the testing page and have ended up with just this error message:
MacOS: unable to find a usable NVRAM partititon - using offset 0x1400.
I am building on Mac OS X, so that could be the problem.
Cormac, the testing page needs minor changes.
The line where configure is set, please change the path from "$HOME/os9/qemu-install" to "$HOME/os9-test/qemu-install". The "-test" part is missing.
After the "Run QEMU with the following command:" text, please remove the "cd ..". That would take the user into the home folder rather than in the os9-test folder.
The command that runs QEMU should be free of any '>' characters. They cause files to be made called -bios, -boot, -ccdrom, -M, and -m. It would probably be easier to copy and paste the command if it were on one line.
At the top of the page it talks about having Python 2 installed. The rest of the directions make no mention of using Python. Are you sure it is needed?
Other than the minor issues, the web site looks good. Thank you.
This patchset significantly improves OpenBIOS's ability to boot OS 9, adding
ESCC legacy ports, reworking ADB interaction with QEMU and fixing IDE
controller interaction with the OS.
Note that the patch to interpreter.fs is a temporary measure pending a fix
that Mark is writing, and that RTAS must be disabled for the OS to boot.
Signed-off-by: Cormac O'Brien <cormac(a)c-obrien.org>
Cormac O'Brien (5):
interpreter.fs: allow evaluate to split words on CR as well as LF
ppc: fix up power-mgt properties
ppc: add ESCC legacy support
ppc: fix CUDA ADB packet header format
ppc: fix up IDE device tree and control transfer
arch/ppc/qemu/methods.c | 2 +
drivers/cuda.c | 17 +++++++--
drivers/escc.c | 85 +++++++++++++++++++++++++++++++++---------
drivers/escc.h | 2 +
drivers/ide.c | 33 ++++++++++++++--
drivers/pci.c | 6 +++
forth/bootstrap/interpreter.fs | 2 +-
7 files changed, 121 insertions(+), 26 deletions(-)
--
2.5.0
