ACPI Testing

Michal Jaegermann michal at harddata.com
Sat Jun 19 14:54:10 UTC 2004


On Fri, Jun 18, 2004 at 11:58:42PM -0400, Douglas Kilpatrick wrote:
> 
> Advice:  If your laptop supports APM, continue to use it.  acpid is much
> more programmable, but since the video won't come back after an ACPI
> suspend/resume cycle unless you are lucky (see the video.txt file mentioned
> above), APM suspend will actually WORK.

The trouble is that you do not have much choice and this depends not
on a kernel but what happens to be in a BIOS of what you have on
hands.  For example my wife's laptop, Acer TM 230, has some
leftovers of APM but if you will try to use it, say to read a
battery status, it bombs in BIOS.  No way to suspend anything with
that.

With ACPI I can get battery just fine and although I had to add
'acpi_sleep=s3_bios' to boot parameters, or video will be not
restored on wakup, and I have to ensure that ehci_hcd module is
unloaded before sleep attempts, which is easy in acpi scripts, but
then it actually suspends, and wakes up, and I do not even have to
do things like 'chvt 1' before sleeping.  The later was a
requirement to prevent crashing video on an old Toshiba with APM.

ACPI looks like a big unwieldy blob which is more often screwed
than not on a firmware level but we may not have much choice in the
matter.  I heard about boxes which will not even boot without ACPI
although I have yet to see one myself.

> Which is probably more important.

Well, yes, but there is no way to predict what will "actually WORK"
on a given machine.

   Michal





More information about the fedora-test-list mailing list