From: jm@jm10.no-ip.com Subject: Asus A6T(c) Date: Sun, 12 Nov 2006 21:34:57 +0100 Message-ID: <45578571.1030905@jm10.no-ip.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: 7bit Return-path: Received: from ble59-2-81-56-123-123.fbx.proxad.net ([81.56.123.123]:30736 "EHLO jm10.no-ip.com") by vger.kernel.org with ESMTP id S1753129AbWKLUfA (ORCPT ); Sun, 12 Nov 2006 15:35:00 -0500 Received: from julien ([192.168.0.2]) by jm10.no-ip.com with esmtpa (Exim 4.50) id 1GjM2c-0005Nv-5F for linux-acpi@vger.kernel.org; Sun, 12 Nov 2006 21:34:58 +0100 Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: linux-acpi@vger.kernel.org Hello, I have Debian/Sid/2.6.18.2-amd64-smp on an Asus A6Tc-AP014H laptop and there are many problems with ACPI. But the most annoying one is that I'd like it to suspend-to-ram. More precisely, it does suspend, but doesn't resume: the hard-drive does not return from power-saving mode. The screen remains off, the system freezes, with solid HD activity light. This is common with SATA disks... but mine is a PATA one. Apart from that, as you asked: > Asus Laptop ACPI Extras version 0.30 > Error calling BSTS > unsupported model A6Tc, trying default values > send /proc/acpi/dsdt to the developers you'll find it on http://81.56.123.123/~jm/A6Tc-AP014H/ with other information about my config. Note there are unknown opcodes, and even after having removed the 4 errors at the beginning of dsdt.dsl, it still fails to recompile. Linux 2.6.18-486 (Debian) boots but Linux 2.6.18-amd64-smp doesn't: it crashes just after > io scheduler noop registered > io scheduler anticipatory registered (default) (or just before "0000:00:0b.0 OHCI: BIOS handoff failed (BIOS bug ?) 000007b4") I need to add either "acpi=noirq" or "noapic" to the kernel command line. "acpi=noirq" gives better results: USB doesn't work with "noapic" (I have exactly the same problem as https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=205672) There are really a lot of awful things in the dmesg: * yet another invitation: > PCI: Bus #04 (-#07) is hidden behind transparent bridge #03 (-#04) (try 'pci=assign-busses') > Please report the result to linux-kernel to fix this permanently * "pcie_portdrv_probe->Dev[02fd:10de] has invalid IRQ. Check vendor BIOS" (I already have the latest BIOS) By curiosity, is CONFIG_ACPI_SLEEP required for suspend-to-ram? Before I enable it, I couldn't write to /sys/power/state: > echo: write error: Operation not permitted Google always tells me CONFIG_HOTPLUG_CPU is required (it was already enabled) but never mentions CONFIG_ACPI_SLEEP, while Linux says: "This option is not recommended for anyone except those doing driver power management development." (drivers/acpi/Kconfig:64) JM