Return-path: Received: from mail-pv0-f174.google.com ([74.125.83.174]:57985 "EHLO mail-pv0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751757Ab0FVPy0 convert rfc822-to-8bit (ORCPT ); Tue, 22 Jun 2010 11:54:26 -0400 Received: by pvh11 with SMTP id 11so438208pvh.19 for ; Tue, 22 Jun 2010 08:54:26 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: References: From: "Luis R. Rodriguez" Date: Tue, 22 Jun 2010 08:54:05 -0700 Message-ID: Subject: Re: ASPM status for iwlwifi and power-management (in general) To: sedat.dilek@gmail.com Cc: wireless , Reinette Chatre , Johannes Berg , John Linville Content-Type: text/plain; charset=UTF-8 Sender: linux-wireless-owner@vger.kernel.org List-ID: On Tue, Jun 22, 2010 at 3:44 AM, Sedat Dilek wrote: > Hi, > > yesterday, I was playing with PowerTop and my iwl3945 still is in the > TOP 3 of energy user. > > Inspired by the recently published ASPM doc from Luis [2] I played a > bit with ASPM. > > Currently, I am using a 2.6.35-rc3 Linux-kernel with latest > acpi-2.6/release fixes. > > # dmesg | grep -i aspm > [    0.226963] pci 0000:01:00.0: disabling ASPM on pre-1.1 PCIe > device.  You can enable it with 'pcie_aspm=force' > [    0.227629] pci 0000:08:00.0: disabling ASPM on pre-1.1 PCIe > device.  You can enable it with 'pcie_aspm=force' > [    0.228205] pci 0000:10:00.0: disabling ASPM on pre-1.1 PCIe > device.  You can enable it with 'pcie_aspm=force' > > # for i in 01:00.0 08:00.0 10:00.0 ; do lspci -v | grep $i ; done > 01:00.0 VGA compatible controller: ATI Technologies Inc M52 [Mobility > Radeon X1300] (prog-if 00 [VGA controller]) > 08:00.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5753M > Gigabit Ethernet PCI Express (rev 21) > 10:00.0 Network controller: Intel Corporation PRO/Wireless 3945ABG > [Golan] Network Connection (rev 02) > > # lspci -vvvv -s 10:00.0 | grep -i aspm >                LnkCap: Port #0, Speed 2.5GT/s, Width x1, ASPM L0s L1, Latency L0 > <128ns, L1 <64us >                LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+ > > Before I will try 'pcie_aspm=force' cheat-code I should note I discourage the usage of pcie_aspm=force, it will try to enable ASPM for all devices which do have the capability, and this means when you are enabling it you are not only testing ASPM on the device you had in mind but *all* devices capable of it. > I wanted to ask about > the status of iwlwifi and ASPM. > Is there code available. yet? Driver code for ASPM only exists to help tweak it if for whatever reason it does need tweaking. An example is ath9k which disables the PLL when it uses ASPM to save power. iwlagn or iwl3945 might have similar stuff, but not sure. Technically drivers should not need any code for ASPM unless its driver specific tweaks. > Which kernel has it - linux-next, > wireless-testing GIT, iwlwifi GIT? ASPM can be used so long as you just have a PCIE bus. If the driver has ASPM tweaks that's a different thing. > Which user-space apps are required? None. > IIRC iw [3] from GIT has already "iw: add set/get power_save commands" [4]. That's different. Read: http://wireless.kernel.org/en/users/Documentation/Power-consumption Luis > Anyone tested ASPM with iwlwifi hardware? Experiences? > > Beyond ASPM code, what is the actual status on power-mangement for > wifi-hardware in general? > IIRC it was turned off by default (with 2.6.33?)? > > Thanks for any help in advance. > > > Kind Regards, > - Sedat - > > > [1] http://wireless.kernel.org/en/users/Documentation/ASPM > [2] http://marc.info/?l=linux-wireless&m=127716115227206&w=2 > [3] http://git.sipsolutions.net/?p=iw.git > [4] http://git.sipsolutions.net/?p=iw.git;a=commit;h=cf40ef379fd6c74a01092d1dfdd936385ea402b0 >