From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 2 Apr 2003 16:11:58 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 2 Apr 2003 16:11:58 -0500 Received: from web11307.mail.yahoo.com ([216.136.131.210]:4624 "HELO web11307.mail.yahoo.com") by vger.kernel.org with SMTP id ; Wed, 2 Apr 2003 16:11:47 -0500 Message-ID: <20030402212311.69823.qmail@web11307.mail.yahoo.com> Date: Wed, 2 Apr 2003 13:23:11 -0800 (PST) From: Subbulu Koya Subject: linux 2.4.20:Intel e1000 4.4.19 and 5.0.43 (Intel Pro1000/MT) hard lock in e1000_reset_hw To: linux-kernel@vger.kernel.org MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org During MAC reset(e1000_hw.c:e1000_reset_hw) my system hard locks. I have tried the following configurations: Software: Linux 2.4.20 SMP Intel e1000 versions 4.4.19 and 5.0.43 with SMP and with and without ANS/diagnostics Hardware: Tyan S2723 Motherboatd w/ dual XEONs *(onboard) Intel 82545EM rev1 (Pro1000/MT) on IRQ 48 (unshared) (also on board) Intel 82557/8/9 (EEPro100) on IRQ 17 (unshared, unplugged) I have traced it to the following point in the 4.4.19 code: e1000_hw.c : e1000_reset_hw(struct e1000_hw *hw) ... /* Issue a global reset to the MAC. This will reset the chip's * transmit, receive, DMA, and link units. It will not effect * the current PCI configuration. The global reset bit is self- * clearing, and should clear within a microsecond. */ DEBUGOUT("Issuing a global reset to MAC\n"); ctrl = E1000_READ_REG(hw, CTRL); if(hw->mac_type > e1000_82543) E1000_WRITE_REG_IO(hw, CTRL, (ctrl | E1000_CTRL_RST)); else E1000_WRITE_REG(hw, CTRL, (ctrl | E1000_CTRL_RST)); ... In this case 'E1000_WRITE_REG_IO(hw, CTRL, (ctrl | E1000_CTRL_RST));' is called. Then the systems continues to msec_delay(4), at which point the system locks up entirely. and in the 5.0.43 code: e1000_main.c: e1000_reset(struct e1000_adapter *adapter) … E1000_WRITE_REG(&adapter->hw, PBA, pba); adapter->hw.fc = adapter->hw.original_fc; e1000_reset_hw(&adapter->hw); … In this case ‘E1000_WRITE_REG(&adapter->hw, PBA, pba);’ is called. Then the system calls e1000_reset_hw, my first trace message isn’t hit and so with this driver version I seem to lockup just a bit earlier then with 4.4.19 In both cases e1000_reset() is being called by e1000_down(). In order to create this problem I bring the interface up and down once every 10seconds or so using ifconfig. Is it possible this driver isn't working properly with the 82545EM rev 1? Is their any additional information I can provide? Thank you in advance, /)dam.. . . D o n ' t S t o p __________________________________________________ Do you Yahoo!? Yahoo! Tax Center - File online, calculators, forms, and more http://tax.yahoo.com