From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263275AbTJQBgg (ORCPT ); Thu, 16 Oct 2003 21:36:36 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263277AbTJQBgg (ORCPT ); Thu, 16 Oct 2003 21:36:36 -0400 Received: from [65.248.22.68] ([65.248.22.68]:36300 "EHLO stoneflynetworks.com") by vger.kernel.org with ESMTP id S263275AbTJQBg1 (ORCPT ); Thu, 16 Oct 2003 21:36:27 -0400 Subject: tg3 driver - trying to get the old v0.99 driver woking on RedHat 2.4.18-10 From: Piet Delaney Reply-To: piet@stoneflynetworks.com To: Linux-Net Cc: Piet Delaney , Piet Delaney , linux-kernel@vger.kernel.org Content-Type: text/plain Organization: StoneFLy Networks Message-Id: <1066354407.2525.40.camel@piet1.stonefly.com> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.2 (1.2.2-4) Date: 16 Oct 2003 18:33:27 -0700 Content-Transfer-Encoding: 7bit X-OriginalArrivalTime: 17 Oct 2003 01:36:26.0606 (UTC) FILETIME=[14A300E0:01C3944F] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I'm curently locked in to using this old RedHat 7.3 with the old 2.4.18 kernel. I need the see what's going on with skbuffs in the ethernet driver, in this case the broadcom bcm5700. When running 2.6 I use the tg3 driver and it's wonderfull. Getting the Broadcom driver to like without being a module looked more difficult that just using an older version of the tg3.c driver. I updated the old 0.99 version (Jun 11, 2002) to know about the Serverworks Card: { PCI_VENDOR_ID_ALTIMA, PCI_DEVICE_ID_ALTIMA_AC1001, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { PCI_VENDOR_ID_ALTIMA, PCI_DEVICE_ID_ALTIMA_AC1003, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { PCI_VENDOR_ID_ALTIMA, PCI_DEVICE_ID_ALTIMA_AC9100, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_0X14, /* Piet */ PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { 0, } and the device is matched ok now. Unfortunately when it does the probe it fails: -------------------------------------------------------------------- (gdb) where #0 pcnet32_probe_pci (pdev=0xc16b3000, ent=0x0) at pcnet32.c:493 #1 0xc028213d in pci_announce_device (drv=, dev=0xc16b3000) at pci.c:626 #2 0xc02821a2 in pci_register_driver (drv=) at pci.c:653 #3 0xc042ffd0 in pcnet32_init_module () at /home/piet2/src/stonefly/perf/eng/ba sebuilds/stonefly/kernel/linux-2.4.18-10/include/linux/pci.h:686 #4 0xc041c7a2 in do_initcalls () at init/main.c:427 #5 0xc010503b in init (unused=0x0) at init/main.c:537 (gdb) ------------------------------------------------------------------------ tg3.c:v0.99 (Jun 11, 2002) tg3: Cannot find proper PCI device base address, aborting. tg3: Cannot find proper PCI device base address, aborting. tg3: Cannot find proper PCI device base address, aborting. PCI: Assigned IRQ 7 for device 02:00.0 tg3: Problem fetching invariants of chip, aborting. PCI: Assigned IRQ 5 for device 02:00.1 tg3: Problem ", aborting. PCI: Assigned IRQ 7 for device 03:06.0 tg3: Problem fetching invariants of chip, aborting. PCI: Assigned IRQ 5 for device 03:06.1 tg3: Problem fetching invariants of chip, aborting. --------------------------------------------------------------------------- What does "fetching invariants of chip" mean? I thought one of you guys might have a suggestion. The tg3 driver looks cleaner than the Brodcom addon. I could also modify the driver so it doesn't have to be built as a module. -piet --