Bug 54622

Summary:	current ACPI implementation problems
Product:	Base System	Reporter:	emil.cazamir <emil.cazamir>
Component:	kern	Assignee:	acpi-jp
Status:	Closed FIXED
Severity:	Affects Only Me
Priority:	Normal
Version:	5.1-RELEASE
Hardware:	Any
OS:	Any

Description emil.cazamir 2003-07-18 22:20:06 UTC

	Current ACPI implementation lacks context save/restore for many
	devices (at least network card and audio card, maybe more) when
	passing through suspend S3 state (STR)

Fix: 

Implementation of an mechanism to save/restore to save device
	context when entering/leaving S3 suspend state.
How-To-Repeat: 	Configure an i386 class machine to use ACPI, set in BIOS setup 
	the suspend type to S3 and execute "acpiconf -s 3". The system 
	goes to STR mode, and when it comes back it leaves many devices in
	an un-usable state. I have encountered problems with the following
	devices: network card (3COM 905 TX-M), sound card (Creative
	SoundBlaster Live! Value), graphics card (ATI Radeon 9500 4P, PN
	113-94232-100; 256bit memory bus, DDRAM). The problems are the
	following:
	  - network becomes unreachable, even if i try ifconfig up/down 
	  - soundcard does not work anymore
	  - the graphics card works almost fine, except cursor (if it was
	    setup to use a text-mode like 80x25 before entering S3) and if 
	    it was setup to use an VESA mode when the sistem comes back from 
	    S3 it shows only the VGA BIOS initialization string (one text
	    mode line at the top of the screen showing "R300 4P Infineon
	    Memory BIOS - P/N 113-94232-100". from this situation I can
	    recover issuing an "vidctl [args]", and only after I enter the
	    command the screen begins to work.

Comment 1 Kris Kennaway freebsd_committer

2003-07-18 23:25:35 UTC

Responsible Changed
From-To: freebsd-bugs->acpi-jp@jp.freebsd.org

Assign to ACPI maintainers

Comment 2 Nate Lawson 2003-07-21 22:57:32 UTC

Unfortunately, the problem is that each driver needs to have proper device
save/restore functions.  This is not the job of ACPI.  Certain devices
currently do work (fxp, sio, pccbb) with ACPI suspend.

Please file separate PRs for each of the devices that isn't working and
post the PRs to the current@ list.  Hopefully the driver maintainers will
implement those functions.

-Nate

Comment 3 mail 2003-07-31 21:01:49 UTC

Hello

The ACPI maintainer sent me a message in which he said that this PR 
should be 'splitted' into few parts, because this problem is not 
related to the ACPI code as it is related to the drivers implementation 
and the way drivers respect ACPI specification. Please close this PR, I 
will send device-related PR's.

Best regards,
Emil Cazamir

Comment 4 Warner Losh freebsd_committer

2003-08-01 23:29:55 UTC

State Changed
From-To: open->closed

Closed per request of the originator.  He'll file multiple followup 
bugs.