From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <2d5a784f0607121826k293267abq63b24333a4f01fa1@mail.gmail.com> Date: Wed, 12 Jul 2006 21:26:05 -0400 From: "david bulkow" To: "Fans of the OS Plan 9 from Bell Labs" <9fans@cse.psu.edu> Subject: Re: Re: [9fans] ACPI In-Reply-To: <44B46AF0.9070901@lanl.gov> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <2d5a784f0607111957kc85c27fmab2f1874e39f5b4a@mail.gmail.com> <44B46AF0.9070901@lanl.gov> Topicbox-Message-UUID: 7e093690-ead1-11e9-9d60-3106f5b1d025 On 7/11/06, Ronald G Minnich wrote: > it would be great if you would (hint). > > And not do it the way linux did it (hint, hint). > I guess I walked into both of these :-) Using the mptable code for clues, I have located and decoded the first couple ACPI tables. Before digging in I was hoping to find someone with which to discuss potential directions. I see this project expanding into three areas: - decode the tables of interest - process AML so I can find most of the tables and eventually handle more than interrupt setup and processor identification - tie all this into the kernel boot sequence As to the first item I seek clarification. Should I follow the example in mp.[ch] and map C structures over each ACPI table or should I decode them in a fashion similar to 9P message conversion? Since this is a rather large project, should I submit code as I complete sections of it, not breaking the kernel of course, or should I package the whole thing up at the end?