Re: PMU, was Re: d-i built (but not tested)
- To: Wouter Verhelst <w@uter.be>
- Cc: Brad Boyer <flar@allandria.com>, debian-68k@lists.debian.org
- Subject: Re: PMU, was Re: d-i built (but not tested)
- From: Finn Thain <fthain@telegraphics.com.au>
- Date: Sat, 10 Sep 2011 22:03:21 +1000 (EST)
- Message-id: <[🔎] alpine.LNX.2.00.1109102148490.2566@nippy.intranet>
- In-reply-to: <alpine.OSX.2.00.1108131722010.189@ibook.local>
- References: <20110726221903.GA30337@grep.be> <alpine.OSX.2.00.1108111233220.189@ibook.local> <20110811135119.GB26179@grep.be> <alpine.OSX.2.00.1108122150490.189@ibook.local> <20110812180108.GA8723@cynthia.pants.nu> <alpine.OSX.2.00.1108131135200.189@ibook.local> <20110813041539.GA11114@cynthia.pants.nu> <20110813065905.GA28584@grep.be> <alpine.OSX.2.00.1108131722010.189@ibook.local>
On Sat, 13 Aug 2011, I wrote:
>
> The PMU driver is started by adb_init() which is a device_initcall, but
> this happens too late. I worked around the problem successfully by
> making it an arch_initcall instead.
This workaround breaks CUDA so I think some kind of solution to the whole
system controller mess is needed.
Even though most PMU machines suffer from a buggy CPU or a broken SCSI
driver or both, this bug still needs to be fixed -- paricularly for
powerbook 190 (which has IDE and no CPU errata) and powerbook 550 (which
has built-in ethernet and no CPU errata).
Finn
Reply to: