Re: 2.6.17-rc5-mm3-lockdep -
On Tue, Jun 06, 2006 at 04:39:21PM +, Stefan Richter wrote:
BTW, the locking in -mm's net/unix/af_unix.c::unix_stream_connect()
differs a bit from stock unix_stream_connect(). I see spin_lock_bh() in
2.6.17-rc5-mm3 where 2.6.17-rc5 has spin_lock().
Hi Ingo:
Looks like this change was introduced by the validator patch. Any idea
why this was done? AF_UNIX is a user-space-driven socket so there shouldn't
be any need for BH to be disabled there.
Even if it does this patch should go through the normal channels rather
than the lock validator.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmVHI~} [EMAIL PROTECTED]
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 3/4] Make Intel e1000 driver legacy I/O port free
Auke Kok wrote:
Kenji Kaneshige wrote:
This patch makes Intel e1000 driver legacy I/O port free.
Signed-off-by: Kenji Kaneshige [EMAIL PROTECTED]
(adding netdev and the other e1000 maintainers to cc:)
without sending this to any of the listed e1000 maintainers *and*
not even including netdev???
I'm sorry about that.
I'm going to take a look at it first, and have some other colleagues
look at this and the impact, which is unclear to me at the moment.
Thank you for reviewing.
What this patch try to do is not to request/enable I/O port regions
if the device can be handled without using I/O port so that the device
can work even if I/O port resource is not assigned to it on the large
servers. Please see the [PATCH 2/4] Update Document/pci.txt about
details.
Thanks,
Kenji Kaneshige
Please don't commit this like this.
Auke
---
drivers/net/e1000/e1000.h |6 +
drivers/net/e1000/e1000_main.c | 130
++---
2 files changed, 75 insertions(+), 61 deletions(-)
Index: linux-2.6.17-rc6/drivers/net/e1000/e1000.h
===
--- linux-2.6.17-rc6.orig/drivers/net/e1000/e1000.h2006-06-06
21:39:11.0 +0900
+++ linux-2.6.17-rc6/drivers/net/e1000/e1000.h2006-06-06
21:56:41.0 +0900
@@ -77,8 +77,9 @@
#define BAR_11
#define BAR_55
-#define INTEL_E1000_ETHERNET_DEVICE(device_id) {\
-PCI_DEVICE(PCI_VENDOR_ID_INTEL, device_id)}
+#define E1000_NO_IOPORT(1 0)
+#define INTEL_E1000_ETHERNET_DEVICE(device_id, flags) {\
+PCI_DEVICE(PCI_VENDOR_ID_INTEL, device_id), .driver_data = flags}
struct e1000_adapter;
@@ -338,6 +339,7 @@
#ifdef NETIF_F_TSO
boolean_t tso_force;
#endif
+int bars;/* BARs to be enabled */
};
Index: linux-2.6.17-rc6/drivers/net/e1000/e1000_main.c
===
--- linux-2.6.17-rc6.orig/drivers/net/e1000/e1000_main.c2006-06-06
21:39:11.0 +0900
+++ linux-2.6.17-rc6/drivers/net/e1000/e1000_main.c2006-06-06
21:56:41.0 +0900
@@ -86,54 +86,54 @@
* {PCI_DEVICE(PCI_VENDOR_ID_INTEL, device_id)}
*/
static struct pci_device_id e1000_pci_tbl[] = {
-INTEL_E1000_ETHERNET_DEVICE(0x1000),
-INTEL_E1000_ETHERNET_DEVICE(0x1001),
-INTEL_E1000_ETHERNET_DEVICE(0x1004),
-INTEL_E1000_ETHERNET_DEVICE(0x1008),
-INTEL_E1000_ETHERNET_DEVICE(0x1009),
-INTEL_E1000_ETHERNET_DEVICE(0x100C),
-INTEL_E1000_ETHERNET_DEVICE(0x100D),
-INTEL_E1000_ETHERNET_DEVICE(0x100E),
-INTEL_E1000_ETHERNET_DEVICE(0x100F),
-INTEL_E1000_ETHERNET_DEVICE(0x1010),
-INTEL_E1000_ETHERNET_DEVICE(0x1011),
-INTEL_E1000_ETHERNET_DEVICE(0x1012),
-INTEL_E1000_ETHERNET_DEVICE(0x1013),
-INTEL_E1000_ETHERNET_DEVICE(0x1014),
-INTEL_E1000_ETHERNET_DEVICE(0x1015),
-INTEL_E1000_ETHERNET_DEVICE(0x1016),
-INTEL_E1000_ETHERNET_DEVICE(0x1017),
-INTEL_E1000_ETHERNET_DEVICE(0x1018),
-INTEL_E1000_ETHERNET_DEVICE(0x1019),
-INTEL_E1000_ETHERNET_DEVICE(0x101A),
-INTEL_E1000_ETHERNET_DEVICE(0x101D),
-INTEL_E1000_ETHERNET_DEVICE(0x101E),
-INTEL_E1000_ETHERNET_DEVICE(0x1026),
-INTEL_E1000_ETHERNET_DEVICE(0x1027),
-INTEL_E1000_ETHERNET_DEVICE(0x1028),
-INTEL_E1000_ETHERNET_DEVICE(0x105E),
-INTEL_E1000_ETHERNET_DEVICE(0x105F),
-INTEL_E1000_ETHERNET_DEVICE(0x1060),
-INTEL_E1000_ETHERNET_DEVICE(0x1075),
-INTEL_E1000_ETHERNET_DEVICE(0x1076),
-INTEL_E1000_ETHERNET_DEVICE(0x1077),
-INTEL_E1000_ETHERNET_DEVICE(0x1078),
-INTEL_E1000_ETHERNET_DEVICE(0x1079),
-INTEL_E1000_ETHERNET_DEVICE(0x107A),
-INTEL_E1000_ETHERNET_DEVICE(0x107B),
-INTEL_E1000_ETHERNET_DEVICE(0x107C),
-INTEL_E1000_ETHERNET_DEVICE(0x107D),
-INTEL_E1000_ETHERNET_DEVICE(0x107E),
-INTEL_E1000_ETHERNET_DEVICE(0x107F),
-INTEL_E1000_ETHERNET_DEVICE(0x108A),
-INTEL_E1000_ETHERNET_DEVICE(0x108B),
-INTEL_E1000_ETHERNET_DEVICE(0x108C),
-INTEL_E1000_ETHERNET_DEVICE(0x1096),
-INTEL_E1000_ETHERNET_DEVICE(0x1098),
-INTEL_E1000_ETHERNET_DEVICE(0x1099),
-INTEL_E1000_ETHERNET_DEVICE(0x109A),
-INTEL_E1000_ETHERNET_DEVICE(0x10B5),
-INTEL_E1000_ETHERNET_DEVICE(0x10B9),
+INTEL_E1000_ETHERNET_DEVICE(0x1000, E1000_NO_IOPORT),
+INTEL_E1000_ETHERNET_DEVICE(0x1001, E1000_NO_IOPORT),
+INTEL_E1000_ETHERNET_DEVICE(0x1004, E1000_NO_IOPORT),
+INTEL_E1000_ETHERNET_DEVICE(0x1008, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x1009, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x100C, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x100D, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x100E, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x100F, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x1010, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x1011, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x1012, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x1013, 0),
+INTEL_E1000_ETHERNET_DEVICE(0x1014, E1000_NO_IOPORT),
+
[PATCH 0/9] NetXen 1G/10G ethernet driver patch with corrections
Hi, I'll be sending a NetXen 1G/10G ethernet driver patch in subsequent emails. We have made changes as per the feedback received. We would like this driver to be inluded in mainline kernel. Kindly review it and feel free to send feedback. Thanks. -Amit Signed-off-by: Amit S. Kale [EMAIL PROTECTED] Kconfig |5 Makefile |1 netxen/Makefile | 35 + netxen/netxen_nic.h | 998 netxen/netxen_nic_ethtool.c | 775 + netxen/netxen_nic_hdr.h | 684 ++ netxen/netxen_nic_hw.c | 1308 +++ netxen/netxen_nic_hw.h | 346 +++ netxen/netxen_nic_init.c | 1263 + netxen/netxen_nic_ioctl.h| 75 ++ netxen/netxen_nic_isr.c | 430 ++ netxen/netxen_nic_main.c | 1228 netxen/netxen_nic_niu.c | 761 + netxen/netxen_nic_phan_reg.h | 293 + 14 files changed, 8202 insertions(+) ~ - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 1/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add Kconfig, Makefile entries and Makefiles for netxen driver
* Add ethtool interface for netxen driver
diff -Naru linux-2.6.16.20.orig/drivers/net/Kconfig
linux-2.6.16.20/drivers/net/Kconfig
--- linux-2.6.16.20.orig/drivers/net/Kconfig2006-06-06 04:48:38.0
-0700
+++ linux-2.6.16.20/drivers/net/Kconfig 2006-06-06 05:02:26.0 -0700
@@ -2313,6 +2313,11 @@
If in doubt, say N.
+config NETXEN_NIC
+ tristate NetXen Multi port (1/10) Gigabit Ethernet NIC
+ help
+ This enables the support for NetXen's Gigabit Ethernet card.
+
endmenu
if !UML
diff -Naru linux-2.6.16.20.orig/drivers/net/Makefile
linux-2.6.16.20/drivers/net/Makefile
--- linux-2.6.16.20.orig/drivers/net/Makefile 2006-06-06 04:48:38.0
-0700
+++ linux-2.6.16.20/drivers/net/Makefile2006-06-06 04:49:32.0
-0700
@@ -213,3 +213,4 @@
obj-$(CONFIG_FS_ENET) += fs_enet/
+obj-$(CONFIG_NETXEN_NIC) += netxen/
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/Makefile
linux-2.6.16.20/drivers/net/netxen/Makefile
--- linux-2.6.16.20.orig/drivers/net/netxen/Makefile1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/Makefile 2006-06-06 06:58:11.0
-0700
@@ -0,0 +1,35 @@
+# Copyright (C) 2003 - 2006 NetXen, Inc.
+# All rights reserved.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License
+# as published by the Free Software Foundation; either version 2
+# of the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+# MA 02111-1307, USA.
+#
+# The full GNU General Public License is included in this distribution
+# in the file called LICENSE.
+#
+# Contact Information:
+#[EMAIL PROTECTED]
+# NetXen,
+# 3965 Freedom Circle, Fourth floor,
+# Santa Clara, CA 95054
+#
+# Makefile for the NetXen NIC Driver
+#
+
+
+obj-$(CONFIG_NETXEN_NIC) := netxen_nic.o
+
+netxen_nic-y := netxen_nic_hw.o netxen_nic_main.o netxen_nic_init.o \
+ netxen_nic_isr.o netxen_nic_ethtool.o netxen_nic_niu.o
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_ethtool.c
linux-2.6.16.20/drivers/net/netxen/netxen_nic_ethtool.c
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_ethtool.c
1969-12-31 16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_ethtool.c 2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,775 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ *
+ *
+ * ethtool support for netxen nic
+ *
+ */
+
+#include linux/types.h
+#include asm/uaccess.h
+#include linux/pci.h
+#include asm/io.h
+#include linux/netdevice.h
+#include linux/ethtool.h
+#include linux/version.h
+
+#include netxen_nic_hw.h
+#include netxen_nic.h
+#include netxen_nic_phan_reg.h
+#include netxen_nic_ioctl.h
+
+struct netxen_nic_stats {
+ char stat_string[ETH_GSTRING_LEN];
+ int sizeof_stat;
+ int stat_offset;
+};
+
+#define NETXEN_NIC_STAT(m) sizeof(((struct netxen_port *)0)-m), \
+offsetof(struct netxen_port, m)
+
+static const struct netxen_nic_stats netxen_nic_gstrings_stats[] = {
+ {rcvd_bad_skb, NETXEN_NIC_STAT(stats.rcvdbadskb)},
+ {xmit_called, NETXEN_NIC_STAT(stats.xmitcalled)},
+ {xmited_frames, NETXEN_NIC_STAT(stats.xmitedframes)},
+ {xmit_finished, NETXEN_NIC_STAT(stats.xmitfinished)},
+ {bad_skb_len, NETXEN_NIC_STAT(stats.badskblen)},
+ {no_cmd_desc, NETXEN_NIC_STAT(stats.nocmddescriptor)},
[PATCH 2/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED] * Add netxen driver main header file diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic.h linux-2.6.16.20/drivers/net/netxen/netxen_nic.h --- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic.h1969-12-31 16:00:00.0 -0800 +++ linux-2.6.16.20/drivers/net/netxen/netxen_nic.h 2006-06-06 06:58:11.0 -0700 @@ -0,0 +1,998 @@ +/* + * Copyright (C) 2003 - 2006 NetXen, Inc. + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, + * MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution + * in the file called LICENSE. + * + * Contact Information: + *[EMAIL PROTECTED] + * NetXen, + * 3965 Freedom Circle, Fourth floor, + * Santa Clara, CA 95054 + */ + +#ifndef _NETXEN_NIC_H_ +#define _NETXEN_NIC_H_ + +#include linux/config.h +#include linux/module.h +#include linux/kernel.h +#include linux/types.h +#include linux/compiler.h +#include linux/slab.h +#include linux/delay.h +#include linux/init.h +#include linux/ioport.h +#include linux/pci.h +#include linux/netdevice.h +#include linux/etherdevice.h +#include linux/ip.h +#include linux/in.h +#include linux/tcp.h +#include linux/skbuff.h +#include linux/version.h + +#include linux/ethtool.h +#include linux/mii.h +#include linux/interrupt.h +#include linux/timer.h + +#include linux/mm.h +#include linux/mman.h + +#include asm/system.h +#include asm/io.h +#include asm/byteorder.h +#include asm/uaccess.h +#include asm/pgtable.h + +#include linux/skbuff.h + +#include netxen_nic_hw.h +#include netxen_nic_hdr.h + +#define NETXEN_NIC_BUILD_NO 231 +#define _NETXEN_NIC_LINUX_MAJOR 2 +#define _NETXEN_NIC_LINUX_MINOR 1 +#define _NETXEN_NIC_LINUX_SUBVERSION 39 +#define NETXEN_NIC_LINUX_VERSIONID 2.1.39 +#define NETXEN_NIC_FW_VERSIONID 2.1.39 +#define NETXEN_NIC_TIMESTAMP Mon May 1 01:34:07 PDT 2006 + +#define RCV_DESC_RINGSIZE(ctxid) \ + (sizeof(struct rcv_desc_t) * adapter-ring_ctx[ctxid].max_rx_desc_count) +#define STATUS_DESC_RINGSIZE(i)\ + (sizeof(struct status_desc_t)* adapter-ring_ctx[i].max_rx_desc_count) +#define TX_RINGSIZE(i) (sizeof(struct netxen_cmd_buffer) \ + * adapter-ring_ctx[i].max_tx_desc_count) +#define JUMBO_DESC_RINGSIZE\ + (sizeof(rcvDesc_t) * adapter-ring_ctx[i].MaxJumboRxDescCount) +#define RCV_BUFFSIZE \ + (sizeof(struct netxen_rx_buffer) * rcv_desc-max_rx_desc_count) + +#define DB_PAGE_SIZE0x1000 +#define NETXEN_CMD_PRODUCER_OFFSET(ctx) \ + ((NETXEN_CMD_PRODUCER_DB_PAGE(ctx)) * DB_PAGE_SIZE) +#define NETXEN_RCV_STATUS_CONSUMER_OFFSET(ctx) \ + ((NETXEN_RCV_STATUS_CONSUMER_DB_PAGE(ctx)) * DB_PAGE_SIZE) +#define NETXEN_RCV_PRODUCER_OFFSET(ctx,ring) \ + ((NETXEN_RCV_PRODUCER_DB_PAGE(ctx,ring)) * DB_PAGE_SIZE) + +#define find_diff_among(a,b,range) ((a)(b)?((b)-(a)):((b)+(range)-(a))) + +#define NETXEN_NETDEV_STATUS 0x1 + +#define ADDR_IN_WINDOW1(off) \ + ((off NETXEN_CRB_PCIX_HOST2) (off NETXEN_CRB_MAX)) ? 1 : 0 + +/* + * normalize a 64MB crb address to 32MB PCI window + * To use NETXEN_CRB_NORMALIZE, window _must_ be set to 1 + */ +#define NETXEN_CRB_NORMALIZE(adapter, reg) \ + (void *)(ptrdiff_t)(adapter-ahw.pci_base+ (reg) \ + - NETXEN_CRB_PCIX_HOST2 + NETXEN_CRB_PCIX_HOST) + +#define DB_NORMALIZE(adapter, off) \ + (void *)(ptrdiff_t)(adapter-ahw.db_base + (off)) + +#define MAX_RX_BUFFER_LENGTH 2000 +#define MAX_RX_JUMBO_BUFFER_LENGTH 9046 +#define RX_DMA_MAP_LEN (MAX_RX_BUFFER_LENGTH - NET_IP_ALIGN) +#define RX_JUMBO_DMA_MAP_LEN \ + (MAX_RX_JUMBO_BUFFER_LENGTH - NET_IP_ALIGN) + +/* + * Maximum number of ring contexts + */ +#define MAX_RING_CTX 1 + +/* Opcodes to be used with the commands */ +#defineTX_ETHER_PKT 0x01 +/* The following opcodes are for IP checksum */ +#defineTX_TCP_PKT 0x02 +#defineTX_UDP_PKT 0x03 +#defineTX_IP_PKT 0x04 +#defineTX_TCP_LSO 0x05 +#defineTX_IPSEC0x06 +#defineTX_IPSEC_CMD0x07 + +/* The following opcodes are for internal consumption. */ +#define NETXEN_CONTROL_OP 0x10 +#define PEGNET_REQUEST 0x11 + +#defineMAX_NUM_CARDS 4 + +#define
[PATCH 3/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED] * Add a header file containing netxen registers info diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_hdr.h linux-2.6.16.20/drivers/net/netxen/netxen_nic_hdr.h --- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_hdr.h1969-12-31 16:00:00.0 -0800 +++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_hdr.h 2006-06-06 06:58:11.0 -0700 @@ -0,0 +1,684 @@ +/* + * Copyright (C) 2003 - 2006 NetXen, Inc. + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, + * MA 02111-1307, USA. + * + * The full GNU General Public License is included in this distribution + * in the file called LICENSE. + * + * Contact Information: + *[EMAIL PROTECTED] + * NetXen, + * 3965 Freedom Circle, Fourth floor, + * Santa Clara, CA 95054 + */ + +#ifndef __NETXEN_NIC_HDR_H_ +#define __NETXEN_NIC_HDR_H_ + +#include linux/module.h +#include linux/kernel.h +#include linux/config.h +#include linux/version.h + +#include asm/semaphore.h +#include linux/spinlock.h +#include asm/irq.h +#include linux/init.h +#include linux/errno.h +#include linux/pci.h +#include linux/types.h +#include asm/uaccess.h +#include asm/string.h /* for memset */ + +extern struct netxen_adapter *g_adapter; + +/* + * The basic unit of access when reading/writing control registers. + */ + +typedef u32 netxen_crbword_t; /* single word in CRB space */ + +#define NETXEN_HW_H0_CH_HUB_ADR0x05 +#define NETXEN_HW_H1_CH_HUB_ADR0x0E +#define NETXEN_HW_H2_CH_HUB_ADR0x03 +#define NETXEN_HW_H3_CH_HUB_ADR0x01 +#define NETXEN_HW_H4_CH_HUB_ADR0x06 +#define NETXEN_HW_H5_CH_HUB_ADR0x07 +#define NETXEN_HW_H6_CH_HUB_ADR0x08 + +/* Hub 0 */ +#define NETXEN_HW_MN_CRB_AGT_ADR 0x15 +#define NETXEN_HW_MS_CRB_AGT_ADR 0x25 + +/* Hub 1 */ +#define NETXEN_HW_PS_CRB_AGT_ADR 0x73 +#define NETXEN_HW_SS_CRB_AGT_ADR 0x20 +#define NETXEN_HW_RPMX3_CRB_AGT_ADR0x0b +#define NETXEN_HW_QMS_CRB_AGT_ADR 0x00 +#define NETXEN_HW_SQGS0_CRB_AGT_ADR0x01 +#define NETXEN_HW_SQGS1_CRB_AGT_ADR0x02 +#define NETXEN_HW_SQGS2_CRB_AGT_ADR0x03 +#define NETXEN_HW_SQGS3_CRB_AGT_ADR0x04 +#define NETXEN_HW_C2C0_CRB_AGT_ADR 0x58 +#define NETXEN_HW_C2C1_CRB_AGT_ADR 0x59 +#define NETXEN_HW_C2C2_CRB_AGT_ADR 0x5a +#define NETXEN_HW_RPMX2_CRB_AGT_ADR0x0a +#define NETXEN_HW_RPMX4_CRB_AGT_ADR0x0c +#define NETXEN_HW_RPMX7_CRB_AGT_ADR0x0f +#define NETXEN_HW_RPMX9_CRB_AGT_ADR0x12 +#define NETXEN_HW_SMB_CRB_AGT_ADR 0x18 + +/* Hub 2 */ +#define NETXEN_HW_NIU_CRB_AGT_ADR 0x31 +#define NETXEN_HW_I2C0_CRB_AGT_ADR 0x19 +#define NETXEN_HW_I2C1_CRB_AGT_ADR 0x29 + +#define NETXEN_HW_SN_CRB_AGT_ADR 0x10 +#define NETXEN_HW_I2Q_CRB_AGT_ADR 0x20 +#define NETXEN_HW_LPC_CRB_AGT_ADR 0x22 +#define NETXEN_HW_ROMUSB_CRB_AGT_ADR 0x21 +#define NETXEN_HW_QM_CRB_AGT_ADR 0x66 +#define NETXEN_HW_SQG0_CRB_AGT_ADR 0x60 +#define NETXEN_HW_SQG1_CRB_AGT_ADR 0x61 +#define NETXEN_HW_SQG2_CRB_AGT_ADR 0x62 +#define NETXEN_HW_SQG3_CRB_AGT_ADR 0x63 +#define NETXEN_HW_RPMX1_CRB_AGT_ADR0x09 +#define NETXEN_HW_RPMX5_CRB_AGT_ADR0x0d +#define NETXEN_HW_RPMX6_CRB_AGT_ADR0x0e +#define NETXEN_HW_RPMX8_CRB_AGT_ADR0x11 + +/* Hub 3 */ +#define NETXEN_HW_PH_CRB_AGT_ADR 0x1A +#define NETXEN_HW_SRE_CRB_AGT_ADR 0x50 +#define NETXEN_HW_EG_CRB_AGT_ADR 0x51 +#define NETXEN_HW_RPMX0_CRB_AGT_ADR0x08 + +/* Hub 4 */ +#define NETXEN_HW_PEGN0_CRB_AGT_ADR0x40 +#define NETXEN_HW_PEGN1_CRB_AGT_ADR0x41 +#define NETXEN_HW_PEGN2_CRB_AGT_ADR0x42 +#define NETXEN_HW_PEGN3_CRB_AGT_ADR0x43 +#define NETXEN_HW_PEGNI_CRB_AGT_ADR0x44 +#define NETXEN_HW_PEGND_CRB_AGT_ADR0x45 +#define NETXEN_HW_PEGNC_CRB_AGT_ADR0x46 +#define NETXEN_HW_PEGR0_CRB_AGT_ADR0x47 +#define NETXEN_HW_PEGR1_CRB_AGT_ADR0x48 +#define NETXEN_HW_PEGR2_CRB_AGT_ADR0x49 +#define NETXEN_HW_PEGR3_CRB_AGT_ADR0x4a + +/* Hub 5 */ +#define NETXEN_HW_PEGS0_CRB_AGT_ADR0x40 +#define NETXEN_HW_PEGS1_CRB_AGT_ADR0x41 +#define NETXEN_HW_PEGS2_CRB_AGT_ADR0x42 +#define NETXEN_HW_PEGS3_CRB_AGT_ADR0x43 +#define NETXEN_HW_PEGSI_CRB_AGT_ADR0x44 +#define NETXEN_HW_PEGSD_CRB_AGT_ADR0x45 +#define
[PATCH 5/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add a header file containing netxen hardware access routines
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_hw.h
linux-2.6.16.20/drivers/net/netxen/netxen_nic_hw.h
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_hw.h 1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_hw.h 2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,346 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ *
+ *
+ * Structures, enums, and macros for the MAC
+ *
+ */
+
+#ifndef __NETXEN_NIC_HW_H_
+#define __NETXEN_NIC_HW_H_
+
+#include netxen_nic_hdr.h
+
+/* Hardware memory size of 128 meg */
+#define NETXEN_MEMADDR_MAX (128 * 1024 * 1024)
+
+#ifndef readq
+static inline u64 readq(void __iomem * addr)
+{
+ return readl(addr) | (((u64) readl(addr + 4)) 32LL);
+}
+#endif
+
+#ifndef writeq
+static inline void writeq(u64 val, void __iomem * addr)
+{
+ writel(((u32) (val)), (addr));
+ writel(((u32) (val 32)), (addr + 4));
+}
+#endif
+
+#define NETXEN_NIC_HW_BLOCK_WRITE_64(DATA_PTR, ADDR, NUM_WORDS) \
+do {\
+int num;\
+u64 *a = (u64 *) (DATA_PTR);\
+u64 *b = (u64 *) (ADDR);\
+for (num = 0; num (NUM_WORDS); num++) { \
+writeq(readq(a), b);\
+b++;\
+a++;\
+} \
+} while (0)
+
+#define NETXEN_NIC_HW_BLOCK_READ_64(DATA_PTR, ADDR, NUM_WORDS)\
+do { \
+int num; \
+u64 *a = (u64 *) (DATA_PTR); \
+u64 *b = (u64 *) (ADDR); \
+for (num = 0; num (NUM_WORDS); num++) { \
+writeq(readq(b), a); \
+b++; \
+a++; \
+} \
+} while (0)
+
+#define NETXEN_PCI_MAPSIZE_BYTES (NETXEN_PCI_MAPSIZE 20)
+#define NETXEN_DB_MAPSIZE_BYTES0x4000
+
+#define NETXEN_NIC_LOCKED_READ_REG(X, Y) \
+addr = (void *)(adapter-ahw.pci_base + X); \
+*(u32 *)Y = readl(addr);
+
+#define NETXEN_NIC_LOCKED_WRITE_REG(X, Y) \
+ addr = (void *)(adapter-ahw.pci_base + X); \
+ writel(*(u32 *)Y, addr);
+
+struct netxen_port;
+void netxen_nic_set_link_parameters(struct netxen_port *port);
+struct netxen_adapter;
+void netxen_nic_flash_print(struct netxen_adapter *adapter);
+
+int netxen_nic_hw_write_wx(struct netxen_adapter *adapter, u64 off,
+ void *data, int len);
+void netxen_crb_writelit_adapter(struct netxen_adapter *adapter,
+unsigned long off, int data);
+int netxen_nic_hw_read_wx(struct netxen_adapter *adapter, u64 off,
+ void *data, int len);
+
+typedef u8 netxen_ethernet_macaddr_t[6];
+
+/* Nibble or Byte mode for phy interface (GbE mode only) */
+typedef enum {
+ NETXEN_NIU_10_100_MB = 0,
+ NETXEN_NIU_1000_MB
+} netxen_niu_gbe_ifmode_t;
+
+/*
+ * NIU GB MAC Config Register 0 (applies to GB0, GB1, GB2, GB3)
+ */
+struct netxen_niu_gb_mac_config_0_t {
+ netxen_crbword_t tx_enable:1, /* 1:enable frame xmit, 0:disable */
+tx_synched:1, /* R/O: xmit enable synched to xmit stream */
+rx_enable:1,
Re: [PATCH 1/9] NetXen 1G/10G ethernet driver patch with corrections
On Wednesday 07 June 2006 11:39, Linsys Contractor Amit S. Kale wrote:
+ switch (mode) {
... You could save a lot of code by using a table and a loop over the registers.
+ case 4:{/* XGB Mode */
+ NETXEN_NIC_LOCKED_READ_REG(NETXEN_NIU_XG_SINGLE_TERM,
+regs_buff[3]);
-Andi
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 4/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add netxen driver hardware access routines
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_hw.c
linux-2.6.16.20/drivers/net/netxen/netxen_nic_hw.c
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_hw.c 1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_hw.c 2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,1308 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ *
+ *
+ * Source file for NIC routines to access the Phantom hardware
+ *
+ */
+
+#include linux/delay.h
+#include linux/netdevice.h
+#include linux/ethtool.h
+#include linux/version.h
+#include netxen_nic.h
+#include netxen_nic_hw.h
+#include netxen_nic_phan_reg.h
+
+/* PCI Windowing for DDR regions. */
+
+#define ADDR_IN_RANGE(addr, low, high) \
+ (((addr) = (high)) ((addr) = (low)))
+
+static unsigned long netxen_nic_pci_set_window(unsigned long long pci_base,
+ unsigned long long addr);
+void netxen_free_hw_resources(struct netxen_adapter *adapter);
+
+int netxen_nic_set_mac(struct net_device *netdev, void *p)
+{
+ struct netxen_port *port = netdev_priv(netdev);
+ struct sockaddr *addr = p;
+#ifdef NETXEN_DEBUG
+ struct netxen_adapter *adapter = port-adapter;
+#endif
+
+ if (netif_running(netdev))
+ return -EBUSY;
+
+ if (!is_valid_ether_addr(addr-sa_data))
+ return -EADDRNOTAVAIL;
+
+ DPRINTK(INFO, valid ether addr\n);
+ memcpy(netdev-dev_addr, addr-sa_data, netdev-addr_len);
+
+ netxen_nic_macaddr_set(port, addr-sa_data);
+
+ return 0;
+}
+
+/**
+ * netxen_nic_set_multi - Multicast
+ **/
+void netxen_nic_set_multi(struct net_device *netdev)
+{
+ struct netxen_port *port = netdev_priv(netdev);
+ struct dev_mc_list *mc_ptr;
+
+ mc_ptr = netdev-mc_list;
+ if (netdev-flags IFF_PROMISC)
+ netxen_nic_set_promisc_mode(port);
+ else
+ netxen_nic_unset_promisc_mode(port);
+}
+
+/*
+ * netxen_nic_change_mtu - Change the Maximum Transfer Unit
+ * @returns 0 on success, negative on failure
+ */
+int netxen_nic_change_mtu(struct net_device *netdev, int new_mtu)
+{
+ struct netxen_port *port = netdev_priv(netdev);
+
+ if (new_mtu 0x)
+ return -EINVAL;
+
+ if (new_mtu 8000) {
+ printk(KERN_ERR %s: %s MTU 8000 is not supported\n,
+ netxen_nic_driver_name, netdev-name);
+ return -EINVAL;
+ }
+
+ netxen_nic_set_mtu(port, new_mtu);
+ netdev-mtu = new_mtu;
+
+ return 0;
+}
+
+/*
+ * check if the firmware has been downloaded and ready to run and
+ * setup the address for the descriptors in the adapter
+ */
+int netxen_nic_hw_resources(struct netxen_adapter *adapter)
+{
+ int i;
+ u32 state = 0;
+ void *addr;
+ int loops = 0, err = 0;
+ int ring;
+ u32 card_cmdring = 0;
+ struct netxen_recv_context *recv_ctx;
+ struct netxen_rcv_desc_ctx *rcv_desc;
+ u32 ring_size;
+ struct netxen_ring_context *ctx;
+
+ DPRINTK(INFO, pci_base: %lx\n, adapter-ahw.pci_base);
+ DPRINTK(INFO, crb_base: %lx %lx, NETXEN_PCI_CRBSPACE,
+ adapter-ahw.pci_base + NETXEN_PCI_CRBSPACE);
+ DPRINTK(INFO, cam base: %lx %lx, NETXEN_CRB_CAM,
+ adapter-ahw.pci_base + NETXEN_CRB_CAM);
+ DPRINTK(INFO, cam RAM: %lx %lx, NETXEN_CAM_RAM_BASE,
+ adapter-ahw.pci_base + NETXEN_CAM_RAM_BASE);
+ DPRINTK(INFO, NIC base:%lx %lx\n, NIC_CRB_BASE_PORT1,
+ adapter-ahw.pci_base + NIC_CRB_BASE_PORT1);
+
+ /* Window 1 call */
+ read_lock(adapter-adapter_lock);
+ card_cmdring = readl(NETXEN_CRB_NORMALIZE(adapter, CRB_CMDPEG_CMDRING));
+ read_unlock(adapter-adapter_lock);
+
+ DPRINTK(INFO, Command Peg sends 0x%x for cmdring base\n,
+
[PATCH 6/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add netxen hardware initialization routines
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_init.c
linux-2.6.16.20/drivers/net/netxen/netxen_nic_init.c
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_init.c 1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_init.c2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,1263 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ *
+ *
+ * Source file for NIC routines to initialize the Phantom Hardware
+ *
+ */
+
+#include linux/netdevice.h
+#include linux/delay.h
+#include netxen_nic.h
+#include netxen_nic_hw.h
+#include netxen_nic_ioctl.h
+#include netxen_nic_phan_reg.h
+
+struct crb_addr_pair {
+ long addr;
+ long data;
+};
+
+#define NETXEN_MAX_CRB_XFORM 60
+static unsigned int crb_addr_xform[NETXEN_MAX_CRB_XFORM];
+#define NETXEN_ADDR_ERROR ((unsigned long ) 0x )
+
+#define crb_addr_transform(name) \
+crb_addr_xform[NETXEN_HW_PX_MAP_CRB_##name] = \
+NETXEN_HW_CRB_HUB_AGT_ADR_##name 20
+
+#define NETXEN_FLASH_BASE (BOOTLD_START)
+#define NETXEN_PHANTOM_MEM_BASE(NETXEN_FLASH_BASE)
+
+static void crb_addr_transform_setup(void)
+{
+ crb_addr_transform(XDMA);
+ crb_addr_transform(TIMR);
+ crb_addr_transform(SRE);
+ crb_addr_transform(SQN3);
+ crb_addr_transform(SQN2);
+ crb_addr_transform(SQN1);
+ crb_addr_transform(SQN0);
+ crb_addr_transform(SQS3);
+ crb_addr_transform(SQS2);
+ crb_addr_transform(SQS1);
+ crb_addr_transform(SQS0);
+ crb_addr_transform(RPMX7);
+ crb_addr_transform(RPMX6);
+ crb_addr_transform(RPMX5);
+ crb_addr_transform(RPMX4);
+ crb_addr_transform(RPMX3);
+ crb_addr_transform(RPMX2);
+ crb_addr_transform(RPMX1);
+ crb_addr_transform(RPMX0);
+ crb_addr_transform(ROMUSB);
+ crb_addr_transform(SN);
+ crb_addr_transform(QMN);
+ crb_addr_transform(QMS);
+ crb_addr_transform(PGNI);
+ crb_addr_transform(PGND);
+ crb_addr_transform(PGN3);
+ crb_addr_transform(PGN2);
+ crb_addr_transform(PGN1);
+ crb_addr_transform(PGN0);
+ crb_addr_transform(PGSI);
+ crb_addr_transform(PGSD);
+ crb_addr_transform(PGS3);
+ crb_addr_transform(PGS2);
+ crb_addr_transform(PGS1);
+ crb_addr_transform(PGS0);
+ crb_addr_transform(PS);
+ crb_addr_transform(PH);
+ crb_addr_transform(NIU);
+ crb_addr_transform(I2Q);
+ crb_addr_transform(EG);
+ crb_addr_transform(MN);
+ crb_addr_transform(MS);
+ crb_addr_transform(CAS2);
+ crb_addr_transform(CAS1);
+ crb_addr_transform(CAS0);
+ crb_addr_transform(CAM);
+ crb_addr_transform(C2C1);
+ crb_addr_transform(C2C0);
+}
+
+static int cmdline_max_cmd_desc = 1024;
+static int cmdline_max_rcv_desc = 16384;
+static int cmdline_max_jumbo_rcv_desc = 512;
+
+module_param(cmdline_max_cmd_desc, int, S_IRUGO);
+module_param(cmdline_max_rcv_desc, int, S_IRUGO);
+module_param(cmdline_max_jumbo_rcv_desc, int, S_IRUGO);
+
+void inline netxen_check_cmdline_params(void)
+{
+ if (cmdline_max_cmd_desc 1024)
+ cmdline_max_cmd_desc = MAX_CMD_DESCRIPTORS;
+ if (cmdline_max_rcv_desc 16384)
+ cmdline_max_rcv_desc = MAX_RCV_DESCRIPTORS;
+ if (cmdline_max_jumbo_rcv_desc 512)
+ cmdline_max_jumbo_rcv_desc = MAX_JUMBO_RCV_DESCRIPTORS;
+}
+
+void __devinit netxen_check_options(struct netxen_adapter *adapter)
+{
+ struct netxen_ring_context *ctx;
+ int i;
+
+ for (i = 0; i MAX_RING_CTX; i++) {
+ ctx = adapter-ring_ctx[i];
+ ctx-max_tx_desc_count = cmdline_max_cmd_desc;
+ ctx-max_rx_desc_count = cmdline_max_rcv_desc;
+ ctx-max_jumbo_rx_desc_count = cmdline_max_jumbo_rcv_desc;
+ }
+}
+
+int
[PATCH 7/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add netxen ioctl interface
* Add netxen interrupt related routines
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_ioctl.h
linux-2.6.16.20/drivers/net/netxen/netxen_nic_ioctl.h
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_ioctl.h 1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_ioctl.h 2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,75 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ */
+
+#ifndef __NETXEN_NIC_IOCTL_H__
+#define __NETXEN_NIC_IOCTL_H__
+
+#include linux/sockios.h
+
+#define NETXEN_CMD_STARTSIOCDEVPRIVATE
+#define NETXEN_NIC_CMD (NETXEN_CMD_START + 1)
+#define NETXEN_NIC_NAME (NETXEN_CMD_START + 2)
+
+typedef enum {
+ netxen_nic_cmd_none = 0,
+ netxen_nic_cmd_pci_read,
+ netxen_nic_cmd_pci_write,
+ netxen_nic_cmd_pci_mem_read,
+ netxen_nic_cmd_pci_mem_write,
+ netxen_nic_cmd_pci_config_read,
+ netxen_nic_cmd_pci_config_write,
+ netxen_nic_cmd_get_stats,
+ netxen_nic_cmd_clear_stats,
+ netxen_nic_cmd_get_version
+} netxen_nic_ioctl_cmd_t;
+
+struct netxen_nic_ioctl_data {
+ u32 cmd;
+ u32 unused1;
+ u64 off;
+ u32 size;
+ u32 rv;
+ char u[64];
+ void *ptr;
+};
+
+struct netxen_statistics {
+ u64 rx_packets;
+ u64 tx_packets;
+ u64 rx_bytes;
+ u64 rx_errors;
+ u64 tx_bytes;
+ u64 tx_errors;
+ u64 rx_crc_errors;
+ u64 rx_short_length_error;
+ u64 rx_long_length_error;
+ u64 rx_mac_errors;
+};
+
+#endif /* __NETXEN_NIC_IOCTL_H_ */
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_isr.c
linux-2.6.16.20/drivers/net/netxen/netxen_nic_isr.c
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_isr.c1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_isr.c 2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,430 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ */
+
+#include linux/netdevice.h
+#include linux/delay.h
+
+#include netxen_nic.h
+#include netxen_nic_hw.h
+#include netxen_nic_phan_reg.h
+
+/*
+ * This will be called when all the ports of the adapter are removed.
+ * This will cleanup and disable interrupts and irq.
+ */
+void netxen_nic_down(struct netxen_port *port)
+{
+ struct netxen_adapter *adapter = port-adapter;
+ int i, j, ctxid;
+ struct netxen_cmd_buffer *cmd_buff;
+ struct netxen_skb_frag *buffrag;
+ struct netxen_ring_context *ctx;
+
+ /* disable phy_ints */
+ netxen_nic_disable_phy_interrupts(adapter, (long)port-portnum);
+
+ adapter-active_ports--;
+
+ if (!adapter-active_ports) {
+ read_lock(adapter-adapter_lock);
+ netxen_nic_disable_int(adapter);
+ read_unlock(adapter-adapter_lock);
+
+ for
[PATCH 8/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add netxen driver main routines: module initialization, probe,
xmit, receive, interrupt handler
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_main.c
linux-2.6.16.20/drivers/net/netxen/netxen_nic_main.c
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_main.c 1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_main.c2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,1228 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ *
+ *
+ * Main source file for NetXen NIC Driver on Linux
+ *
+ */
+
+#include netxen_nic.h
+
+#define DEFINE_GLOBAL_RECV_CRB
+#include netxen_nic_phan_reg.h
+#include netxen_nic_ioctl.h
+
+MODULE_DESCRIPTION(NetXen Multi port (1/10) Gigabit Network Driver);
+MODULE_LICENSE(GPL);
+MODULE_VERSION(NETXEN_NIC_LINUX_VERSIONID);
+
+char netxen_nic_driver_name[] = netxen;
+char netxen_nic_driver_string[] = NetXen Network Driver version
+NETXEN_NIC_LINUX_VERSIONID
+- NETXEN_NIC_BUILD_NO generated NETXEN_NIC_TIMESTAMP;
+
+#define NETXEN_NETDEV_WEIGHT 120
+#define NETXEN_ADAPTER_UP_MAGIC 777
+
+static int netxen_nic_probe_err = 0;
+
+/* Local functions to NetXen NIC driver */
+static int __devinit netxen_nic_probe(struct pci_dev *pdev,
+ const struct pci_device_id *ent);
+static void __devexit netxen_nic_remove(struct pci_dev *pdev);
+static int netxen_nic_open(struct net_device *netdev);
+static int netxen_nic_close(struct net_device *netdev);
+static int netxen_nic_xmit_frame(struct sk_buff *, struct net_device *);
+static void netxen_tx_timeout(struct net_device *netdev);
+static void netxen_tx_timeout_task(struct net_device *netdev);
+static void netxen_watchdog(unsigned long);
+static int netxen_handle_int(struct netxen_adapter *, struct net_device *);
+static int netxen_nic_ioctl(struct net_device *netdev,
+ struct ifreq *ifr, int cmd);
+static int netxen_nic_poll(struct net_device *dev, int *budget);
+#ifdef CONFIG_NET_POLL_CONTROLLER
+static void netxen_nic_poll_controller(struct net_device *netdev);
+#endif
+static irqreturn_t netxen_intr(int irq, void *data, struct pt_regs *regs);
+
+/* PCI Device ID Table */
+static struct pci_device_id netxen_pci_tbl[] __devinitdata = {
+ NETXEN_PCI_ID(PCI_DEVICE_ID_NX_QG),
+ NETXEN_PCI_ID(PCI_DEVICE_ID_NX_XG),
+ NETXEN_PCI_ID(PCI_DEVICE_ID_NX_CX4),
+ NETXEN_PCI_ID(PCI_DEVICE_ID_NX_IMEZ),
+ NETXEN_PCI_ID(PCI_DEVICE_ID_NX_HMEZ),
+ {0,}
+};
+
+MODULE_DEVICE_TABLE(pci, netxen_pci_tbl);
+
+struct netxen_adapter *adapterlist[MAX_NUM_CARDS];
+
+/*
+ * netxen_nic_probe()
+ *
+ * Linux system will invoke this after identifying the vendor ID and device Id
+ * in the pci_tbl where this module will search for NetXen vendor and device
ID for
+ * quad port adapter.
+ *
+ * Even though there are 4 seperate devices (functions) on the quad port NIC,
but
+ * the probe will be invoked for the first device (function) only. This will
+ * initialize the adapter, and setup the global parameters along with the
port's
+ * specific structure.
+ */
+static int __devinit
+netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+ struct net_device *netdev = NULL;
+ struct netxen_adapter *adapter = NULL;
+ struct netxen_port *port = NULL;
+ u8 *mem_ptr = NULL;
+ u8 *db_ptr = NULL;
+ unsigned long mem_base, mem_len, db_base, db_len;
+ int pci_using_dac, i, err;
+ int boardno = 0;
+ int ring;
+ struct netxen_recv_context *recv_ctx = NULL;
+ struct netxen_rcv_desc_ctx *rcv_desc = NULL;
+ struct netxen_cmd_buffer *cmd_buf_arr[MAX_RING_CTX] = { NULL };
+ struct netxen_ring_context *ctx;
+ static int netxen_cards_found = 0;
+ u64 mac_addr[FLASH_NUM_PORTS + 1];
+ int valid_mac;
+
+ if ((err = pci_enable_device(pdev))) {
+
[PATCH 9/9] NetXen 1G/10G ethernet driver patch with corrections
2006-06-07 Amit S. Kale [EMAIL PROTECTED]
* Add netxen driver niu (hardware phy interface component)
handling routines
* Add netxen hardware CRB register definitions
diff -Naru linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_niu.c
linux-2.6.16.20/drivers/net/netxen/netxen_nic_niu.c
--- linux-2.6.16.20.orig/drivers/net/netxen/netxen_nic_niu.c1969-12-31
16:00:00.0 -0800
+++ linux-2.6.16.20/drivers/net/netxen/netxen_nic_niu.c 2006-06-06
06:58:11.0 -0700
@@ -0,0 +1,761 @@
+/*
+ * Copyright (C) 2003 - 2006 NetXen, Inc.
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.
+ *
+ * Contact Information:
+ *[EMAIL PROTECTED]
+ * NetXen,
+ * 3965 Freedom Circle, Fourth floor,
+ * Santa Clara, CA 95054
+ *
+ *
+ * Provides access to the Network Interface Unit h/w block.
+ *
+ */
+
+#include netxen_nic.h
+#include linux/delay.h
+
+/**
+ * netxen_niu_gbe_phy_read - read a register from the GbE PHY via
+ * mii management interface.
+ *
+ * Note: The MII management interface goes through port 0.
+ * Individual phys are addressed as follows:
+ * @param phy [15:8] phy id
+ * @param reg [7:0] register number
+ *
+ * @returns 0 on success
+ * -1 on error
+ *
+ **/
+long netxen_niu_gbe_phy_read(struct netxen_adapter *adapter, long phy,
+long reg, netxen_crbword_t * readval)
+{
+ long timeout = 0;
+ long result = 0;
+ long restore = 0;
+ struct netxen_niu_gb_mii_mgmt_address address;
+ struct netxen_niu_gb_mii_mgmt_command command;
+ struct netxen_niu_gb_mii_mgmt_indicators status;
+ struct netxen_niu_gb_mii_mgmt_config mii_cfg;
+ struct netxen_niu_gb_mac_config_0_t mac_cfg0;
+
+ /* MII mgmt all goes through port 0 MAC interface, so it cannot be in
reset */
+ if (netxen_nic_hw_read_wx(adapter, NETXEN_NIU_GB_MAC_CONFIG_0(0),
+ mac_cfg0, 4))
+ return -EIO;
+ if (mac_cfg0.soft_reset) {
+ struct netxen_niu_gb_mac_config_0_t temp;
+ *(netxen_crbword_t *) temp = 0;
+ temp.tx_reset_pb = 1;
+ temp.rx_reset_pb = 1;
+ temp.tx_reset_mac = 1;
+ temp.rx_reset_mac = 1;
+ if (netxen_nic_hw_write_wx(adapter,
+ NETXEN_NIU_GB_MAC_CONFIG_0(0),
+ temp, 4))
+ return -EIO;
+ restore = 1;
+ }
+
+ /* reset MII management interface */
+ *(netxen_crbword_t *) mii_cfg = 0;
+ mii_cfg.clockselect = 7;
+ mii_cfg.reset = 1;
+ if (netxen_nic_hw_write_wx(adapter, NETXEN_NIU_GB_MII_MGMT_CONFIG(0),
+ mii_cfg, 4))
+ return -EIO;
+ mii_cfg.reset = 0;
+ if (netxen_nic_hw_write_wx(adapter, NETXEN_NIU_GB_MII_MGMT_CONFIG(0),
+ mii_cfg, 4))
+ return -EIO;
+
+ *(netxen_crbword_t *) address = 0;
+ address.reg_addr = reg;
+ address.phy_addr = phy;
+ if (netxen_nic_hw_write_wx(adapter, NETXEN_NIU_GB_MII_MGMT_ADDR(0),
+ address, 4))
+ return -EIO;
+ *(netxen_crbword_t *) command = 0;/* turn off any prior
activity */
+ if (netxen_nic_hw_write_wx(adapter, NETXEN_NIU_GB_MII_MGMT_COMMAND(0),
+ command, 4))
+ return -EIO;
+ /* send read command */
+ command.read_cycle = 1;
+ if (netxen_nic_hw_write_wx(adapter, NETXEN_NIU_GB_MII_MGMT_COMMAND(0),
+ command, 4))
+ return -EIO;
+
+ *(netxen_crbword_t *) status = 0;
+ do {
+ if (netxen_nic_hw_read_wx(adapter,
+ NETXEN_NIU_GB_MII_MGMT_INDICATE(0),
+ status, 4))
+ return -EIO;
+ timeout++;
+ } while ((status.busy || status.notvalid)
+ (timeout++ NETXEN_NIU_PHY_WAITMAX));
+
+ if
Re: [PATCH] Right prototype of __raw_v4_lookup()
On Mon, Jun 05, 2006 at 09:07:07PM -0700, David Miller wrote: From: Alexey Dobriyan [EMAIL PROTECTED] Date: Sun, 28 May 2006 01:09:32 +0400 All users pass 32-bit values as addresses and internally they're compared with 32-bit entities. So, change laddr and raddr types to __be32. Applied. Doesn't this mean that, at least eventually, we should change the address members of inet_sock to __be32 as well? I think, yes. Al Viro is sitting on terabytes of endian annotations in networking code. See net-endian.b* branches at git://git.kernel.org/pub/scm/linux/kernel/git/viro/bird.git I don't know if he considers them ready. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
On Tue, 2006-06-06 at 14:24 -0500, Larry Finger wrote: Jun 6 13:34:10 larrylap kernel: SoftMAC: generic IE set to ffdd160050fff2010150fff2020150fff2020150fff202 Why does SoftMAC set a generic IE rather than get one from wpa_supplicant? I would appreciate help with this problem. generic IE is what the wext is called, it is the IE that wpa_supplicant sets. Not sure what's going on though, I know next to nothing about wpa. johannes signature.asc Description: This is a digitally signed message part
[patch] workaround zd1201 interference problem
zd1201 likes to start up shouting, interfering with all wifis in range. It is capable of stopping ipw2200 up-to few meters away, and stops other cards at smaller distances, too. This works around it. Only forcing ZD1201_CMDCODE_DISABLE is not enough to prevent interference. From: Jirka Benc [EMAIL PROTECTED] Signed-off-by: Pavel Machek [EMAIL PROTECTED] --- linux-good/drivers/net/wireless/zd1201.c2006-03-30 13:51:58.0 +0200 +++ linux/drivers/net/wireless/zd1201.c 2006-06-07 15:55:01.0 +0200 @@ -1835,6 +1835,8 @@ zd-dev-name); usb_set_intfdata(interface, zd); + zd1201_enable(zd); /* zd1201 likes to startup shouting, interfering */ + zd1201_disable(zd); /* with all the wifis in range */ return 0; err_net: -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [patch] workaround zd1201 interference problem
On Wed, 7 Jun 2006 16:00:45 +0200, Pavel Machek wrote: + zd1201_enable(zd); /* zd1201 likes to startup shouting, interfering */ + zd1201_disable(zd); /* with all the wifis in range */ I would prefer to track it down and find out where exactly is the problem instead of this quick hack. Jiri -- Jiri Benc SUSE Labs - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [patch] workaround zd1201 interference problem
On St 07-06-06 16:08:28, Jiri Benc wrote: On Wed, 7 Jun 2006 16:00:45 +0200, Pavel Machek wrote: + zd1201_enable(zd); /* zd1201 likes to startup shouting, interfering */ + zd1201_disable(zd); /* with all the wifis in range */ I would prefer to track it down and find out where exactly is the problem instead of this quick hack. Well, I'll try _enable() alone, but it seems to me that _enable() command is needed to initialize radio properly. I do not think we can get much further without firmware sources... -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 3/4] Make Intel e1000 driver legacy I/O port free
Kenji Kaneshige wrote: Auke Kok wrote: Kenji Kaneshige wrote: This patch makes Intel e1000 driver legacy I/O port free. Signed-off-by: Kenji Kaneshige [EMAIL PROTECTED] (adding netdev and the other e1000 maintainers to cc:) without sending this to any of the listed e1000 maintainers *and* not even including netdev??? I'm sorry about that. I also didn't see that you were sending this to Greg-KH. I think I got thrown off by that as I wasn't following lkml until yesterday in the first place. I'll toss the patches around over here and see what comes up. Cheers, Auke - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] e1000: fix netpoll with NAPI
Matt, any ideas on this?
Not at the moment.
how about this for a solution? It doesn't make netpoll any more robust, but I
think in the interests of efficiency it would be fair to require that, when
netpolled, a driver must receive frames on the same net device for which it was
polled. With this patch we detect that condition and handle it accordingly in
e1000_intr. This eliminates the need for us to call the clean_rx method from
the poll_controller when napi is configured, instead allowing the poll method to
be called from napi_poll, as the netpoll model currently does. This fixes the
netdump regression, and eliminates the layering violation and the potential race
that we've been discussing. I've just tested it with netdump here and it works
quite well.
Thoughts appreciated.
Thanks Regards
Neil
Signed-off-by: Neil Horman [EMAIL PROTECTED]
e1000_main.c | 54 +-
1 files changed, 45 insertions(+), 9 deletions(-)
--- linux-2.6.9/drivers/net/e1000/e1000_main.c.neil 2006-06-06
10:37:42.0 -0400
+++ linux-2.6.9/drivers/net/e1000/e1000_main.c 2006-06-07 10:48:22.0
-0400
@@ -3207,8 +3207,9 @@
* @pt_regs: CPU registers structure
**/
+
static irqreturn_t
-e1000_intr(int irq, void *data, struct pt_regs *regs)
+__e1000_intr(int irq, void *data, struct pt_regs *regs, int netpoll_op)
{
struct net_device *netdev = data;
struct e1000_adapter *adapter = netdev_priv(netdev);
@@ -3217,6 +3218,7 @@
#ifndef CONFIG_E1000_NAPI
int i;
#else
+ struct net_device *dev_to_sched;
/* Interrupt Auto-Mask...upon reading ICR,
* interrupts are masked. No need for the
* IMC write, but it does mean we should
@@ -3255,8 +3257,22 @@
E1000_WRITE_REG(hw, IMC, ~0);
E1000_WRITE_FLUSH(hw);
}
- if (likely(netif_rx_schedule_prep(adapter-polling_netdev[0])))
- __netif_rx_schedule(adapter-polling_netdev[0]);
+
+ /*
+* netpoll operations, in the interests of efficiency
+* only do napi polling on the device passed to the
+* poll_controller. Therefore, if we are preforming
+* a netpoll operation, then we can't schedule a receive
+* to one of the dummy net devices that exist for sole
+* purpose of spreading out rx schedules
+*/
+ if (netpoll_op)
+ dev_to_sched = netdev;
+ else
+ dev_to_sched = adapter-polling_netdev[0];
+
+ if (likely(netif_rx_schedule_prep(dev_to_sched)))
+ __netif_rx_schedule(dev_to_sched);
else
e1000_irq_enable(adapter);
#else
@@ -3288,6 +3304,13 @@
return IRQ_HANDLED;
}
+static irqreturn_t
+e1000_intr(int irq, void *data, struct pt_regs *regs)
+{
+ return __e1000_intr(irq, data, regs, 0);
+}
+
+
#ifdef CONFIG_E1000_NAPI
/**
* e1000_clean - NAPI Rx polling callback
@@ -3300,7 +3323,6 @@
struct e1000_adapter *adapter;
int work_to_do = min(*budget, poll_dev-quota);
int tx_cleaned = 0, i = 0, work_done = 0;
-
/* Must NOT use netdev_priv macro here. */
adapter = poll_dev-priv;
@@ -3308,10 +3330,24 @@
if (!netif_carrier_ok(adapter-netdev))
goto quit_polling;
- while (poll_dev != adapter-polling_netdev[i]) {
- i++;
- if (unlikely(i == adapter-num_rx_queues))
- BUG();
+ /*
+* only search for a matching polling_netdev in the event
+* that this isn't a real registered net_device
+* A real net device can be passed in here in the event
+* that netdump has been activated (this comes through
+* netpoll_poll_dev). We detect this by virtue of the
+* fact that each polling_netdev-priv points to the private
+* data of its parent (registered) netdev. So if:
+* poll_dev-priv == netdev_priv(poll_dev), its a real device
+* otherwise its a polling_netdev.
+*/
+ if (adapter != netdev_priv(poll_dev)) {
+ while (poll_dev != adapter-polling_netdev[i]) {
+ i++;
+ if (unlikely(i == adapter-num_rx_queues))
+ BUG();
+ }
+
}
if (likely(adapter-num_tx_queues == 1)) {
@@ -4624,7 +4660,7 @@
{
struct e1000_adapter *adapter = netdev_priv(netdev);
disable_irq(adapter-pdev-irq);
- e1000_intr(adapter-pdev-irq, netdev, NULL);
+ __e1000_intr(adapter-pdev-irq, netdev, NULL, 1);
e1000_clean_tx_irq(adapter, adapter-tx_ring);
#ifndef CONFIG_E1000_NAPI
adapter-clean_rx(adapter, adapter-rx_ring);
--
/***
*Neil Horman
*Software Engineer
*gpg keyid: 1024D / 0x92A74FA1 - http://pgp.mit.edu
***/
-
To unsubscribe
[PATCH 0/4] ehea: IBM eHEA Ethernet Device Driver
Hello, this is the first version of the IBM eHEA Ethernet Device Driver. It supports a new IBM ethernet chip on System p. Main functionality including broadcast multicast and some parts of ethtool already work. Things we're still working on: - vlan support - performance improvements in SMP systems - kernel coding style This code is not ready for kernel inclusion as is, but we definetly want to get there. The patch series consists of 4 patches 01: interfaces to network stack and headers 02: pHYP interface 03: queue managment 04: Kconfig and Makefile Signed-off-by: Jan-Bernd Themann [EMAIL PROTECTED] drivers/net/ehea/Kconfig |6 drivers/net/ehea/Makefile| 45 drivers/net/ehea/ehea.h | 434 +++ drivers/net/ehea/ehea_hw.h | 319 + drivers/net/ehea/ehea_main.c | 2571 +++ drivers/net/ehea/ehea_phyp.c | 1026 + drivers/net/ehea/ehea_phyp.h | 625 ++ drivers/net/ehea/ehea_qmr.c | 719 drivers/net/ehea/ehea_qmr.h | 390 ++ 9 files changed, 6135 insertions(+) - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
Johannes Berg wrote: generic IE is what the wext is called, it is the IE that wpa_supplicant sets. Not sure what's going on though, I know next to nothing about wpa. I have a little more information on what is happening. In IEEE Std 802.11i-2004, which defines the WPA protocol, Figure 11a shows the sequence of exchanges needed to associate. Both bcm43xx-softmac and ndiswrapper go through the Open System Authentication process. Where they seem to diverge is in the STA's Association Request (Security Parameters) step. With ndiswrapper, the AP responds with a WPA EAPOL-Key message; whereas with softmac, the AP sends back the invalid pairwise cipher message and rejects the association. Can anyone point to a reference that states what the content of the Association Request should be to get the AP to respond with the EAPOL-Key message? Unfortunately, I have no possibility of implementing a sniffer to see what a correct message contains. Larry - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
HELP! Slip Interface losing packets. Looking for contractor to fix.
Hello, We have a custom compiled Linux Kernel running oh RHEL 3.1 [EMAIL PROTECTED] root]# uname -a Linux st0056_1 2.4.21-32.0.1.ELcustom #12 SMP Wed Sep 14 11:55:22 EDT 2005 i686 i686 i386 GNU/Linux We compiled in SLIP support using the standard RHEL kernel supplied by Redhat. Unfortunately, RH does not support SLIP. We are experiencing problems that we believe are a kernel issue (either SLIP or SERIAL) and are interested in hiring a contractor to diagnose the problem and provide a kernel patch to the kernel we are using (if necessary). We do not want to upgrade or change distributions at this time (business/practicality reasons). If you are interested, please send an e-mail with your resume (doc, odf, or plain text) to '[EMAIL PROTECTED]' (I'll provide a private e-mail address to interested parties). The company website is here: http://www.sterlingjewelers.com/ Thank You, Gerald Butler Project Manager - Store Systems Sterling Inc The information contained in this e-mail message is privileged and/or confidential and is intended only for the use of the individual or entity named above. If the reader of this message is not the intended recipient, or the employee or agent responsible to deliver it to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please immediately notify us by telephone (330-668-5000), and destroy the original message. Thank you. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
On Wed, 2006-06-07 at 10:47 -0500, Larry Finger wrote: I have a little more information on what is happening. Great. In IEEE Std 802.11i-2004, which defines the WPA protocol, Figure 11a shows the sequence of exchanges needed to associate. Both bcm43xx-softmac and ndiswrapper go through the Open System Authentication process. Right, you always have to do that. Where they seem to diverge is in the STA's Association Request (Security Parameters) step. With ndiswrapper, the AP responds with a WPA EAPOL-Key message; whereas with softmac, the AP sends back the invalid pairwise cipher message and rejects the association. Interesting. That's strange. Can anyone point to a reference that states what the content of the Association Request should be to get the AP to respond with the EAPOL-Key message? Unfortunately, I have no possibility of implementing a sniffer to see what a correct message contains. Well, it should be shown in the 802.11i spec too. johannes signature.asc Description: This is a digitally signed message part
Re: Problem authenticating using WPA with bcm43xx-softmac
On Wed, 2006-06-07 at 17:51 +0200, Johannes Berg wrote: Well, it should be shown in the 802.11i spec too. I suppose that it is the association request, and needs to contain the RSN described in 7.3.2.25 as per 7.2.3.4 in 802.11i. This is, afaik, the 'generic IE' that is added with the wext. Now, it looks like the RSN isn't included but the WPA2 info or something? Also, the genIE in your log doesn't look correct to me, starting with ff?? Jouni, do you have any idea what might be going on? johannes signature.asc Description: This is a digitally signed message part
Re: Problem authenticating using WPA with bcm43xx-softmac
Johannes Berg wrote: On Wed, 2006-06-07 at 10:47 -0500, Larry Finger wrote: I have a little more information on what is happening. Great. In IEEE Std 802.11i-2004, which defines the WPA protocol, Figure 11a shows the sequence of exchanges needed to associate. Both bcm43xx-softmac and ndiswrapper go through the Open System Authentication process. Right, you always have to do that. Where they seem to diverge is in the STA's Association Request (Security Parameters) step. With ndiswrapper, the AP responds with a WPA EAPOL-Key message; whereas with softmac, the AP sends back the invalid pairwise cipher message and rejects the association. Interesting. That's strange. Can anyone point to a reference that states what the content of the Association Request should be to get the AP to respond with the EAPOL-Key message? Unfortunately, I have no possibility of implementing a sniffer to see what a correct message contains. Well, it should be shown in the 802.11i spec too. Beware of the order of IE's in the management frames; some AP's are touchy about this. Sam - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
On Wed, 2006-06-07 at 09:01 -0700, Sam Leffler wrote: Beware of the order of IE's in the management frames; some AP's are touchy about this. Uh oh. I have no idea where the ieee80211 layer sticks that one, probably at the end. johannes signature.asc Description: This is a digitally signed message part
Re: Problem authenticating using WPA with bcm43xx-softmac
Sam Leffler wrote: Beware of the order of IE's in the management frames; some AP's are touchy about this. That may be the cause here as the behavior when I changed AP's from the Linux-based model to one with VXWorks. Larry - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
Johannes Berg wrote: On Wed, 2006-06-07 at 09:01 -0700, Sam Leffler wrote: Beware of the order of IE's in the management frames; some AP's are touchy about this. Uh oh. I have no idea where the ieee80211 layer sticks that one, probably at the end. I can verify that. I'm currently building a new module with a printk to verify that it is being added. If it is, I'll try moving it to see if it helps. Larry - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 2/7] AMSO1100 Low Level Driver.
On Wed, 2006-05-31 at 11:59 -0700, Stephen Hemminger wrote:
The following should be replaced with BUG_ON() or WARN_ON().
and pr_debug()
+#ifdef C2_DEBUG
+#define assert(expr) \
+if(!(expr)) { \
+printk(KERN_ERR PFX Assertion failed! %s, %s, %s, line %d\n,\
+ #expr, __FILE__, __FUNCTION__, __LINE__); \
+}
+#define dprintk(fmt, args...) do {printk(KERN_INFO PFX fmt, ##args);} while
(0)
+#else
+#define assert(expr) do {} while (0)
+#define dprintk(fmt, args...) do {} while (0)
+#endif /* C2_DEBUG */
Also, you tend to use assert() as a bogus NULL pointer check.
If you get passed a NULL, it is a bug, and the deref will fail
and cause a pretty stack dump...
done.
+static void c2_set_rxbufsize(struct c2_port *c2_port)
+{
+ struct net_device *netdev = c2_port-netdev;
+
+ assert(netdev != NULL);
Bogus, you will just fail on the deref below
done.
+
+ if (netdev-mtu RX_BUF_SIZE)
+ c2_port-rx_buf_size =
+ netdev-mtu + ETH_HLEN + sizeof(struct c2_rxp_hdr) +
+ NET_IP_ALIGN;
+ else
+ c2_port-rx_buf_size = sizeof(struct c2_rxp_hdr) + RX_BUF_SIZE;
+}
+static void c2_rx_interrupt(struct net_device *netdev)
+{
+ struct c2_port *c2_port = netdev_priv(netdev);
+ struct c2_dev *c2dev = c2_port-c2dev;
+ struct c2_ring *rx_ring = c2_port-rx_ring;
+ struct c2_element *elem;
+ struct c2_rx_desc *rx_desc;
+ struct c2_rxp_hdr *rxp_hdr;
+ struct sk_buff *skb;
+ dma_addr_t mapaddr;
+ u32 maplen, buflen;
+ unsigned long flags;
+
+ spin_lock_irqsave(c2dev-lock, flags);
+
+ /* Begin where we left off */
+ rx_ring-to_clean = rx_ring-start + c2dev-cur_rx;
+
+ for (elem = rx_ring-to_clean; elem-next != rx_ring-to_clean;
+ elem = elem-next) {
+ rx_desc = elem-ht_desc;
+ mapaddr = elem-mapaddr;
+ maplen = elem-maplen;
+ skb = elem-skb;
+ rxp_hdr = (struct c2_rxp_hdr *) skb-data;
+
+ if (rxp_hdr-flags != RXP_HRXD_DONE)
+ break;
+ buflen = rxp_hdr-len;
+
+ /* Sanity check the RXP header */
+ if (rxp_hdr-status != RXP_HRXD_OK ||
+ buflen (rx_desc-len - sizeof(*rxp_hdr))) {
+ c2_rx_error(c2_port, elem);
+ continue;
+ }
+
+ /*
+ * Allocate and map a new skb for replenishing the host
+ * RX desc
+ */
+ if (c2_rx_alloc(c2_port, elem)) {
+ c2_rx_error(c2_port, elem);
+ continue;
+ }
+
+ /* Unmap the old skb */
+ pci_unmap_single(c2dev-pcidev, mapaddr, maplen,
+ PCI_DMA_FROMDEVICE);
+
prefetch(skb-data) here will help performance.
good. ok.
+ /*
+ * Skip past the leading 8 bytes comprising of the
+ * struct c2_rxp_hdr, prepended by the adapter
+ * to the usual Ethernet header (struct ethhdr),
+ * to the start of the raw Ethernet packet.
+ *
+ * Fix up the various fields in the sk_buff before
+ * passing it up to netif_rx(). The transfer size
+ * (in bytes) specified by the adapter len field of
+ * the struct rxp_hdr_t does NOT include the
+ * sizeof(struct c2_rxp_hdr).
+ */
+ skb-data += sizeof(*rxp_hdr);
+ skb-tail = skb-data + buflen;
+ skb-len = buflen;
+ skb-dev = netdev;
+ skb-protocol = eth_type_trans(skb, netdev);
+
+ /* Drop arp requests to the pseudo nic ip addr */
+ if (unlikely(ntohs(skb-protocol) == ETH_P_ARP)) {
+ u8 *tpa;
+
+ /* pull out the tgt ip addr */
+ tpa = skb-data /* beginning of the arp packet */
+ + 8 /* arp addr fmts, lens, and opcode */
+ + 6 /* arp src hw addr */
+ + 4 /* arp src proto addr */
+ + 6;/* arp tgt hw addr */
+ if (is_rnic_addr(c2dev-pseudo_netdev, *((u32 *)tpa))) {
+ dprintk(Dropping arp req for
+ %03d.%03d.%03d.%03d\n,
+ tpa[0], tpa[1], tpa[2], tpa[3]);
+ kfree_skb(skb);
+ continue;
+ }
+ }
This is looks like a mess, please do it
Re: [patch] workaround zd1201 interference problem
Pavel Machek wrote: Well, I'll try _enable() alone, but it seems to me that _enable() command is needed to initialize radio properly. I do not think we can get much further without firmware sources... If you can formulate a proper and technical description of the issue (and exactly what is needed to workaround it), I can contact ZyDAS for you. They have been very helpful with ZD1211. Daniel - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] e1000: fix netpoll with NAPI
On Wed, Jun 07, 2006 at 11:05:22AM -0400, Neil Horman wrote: Matt, any ideas on this? Not at the moment. how about this for a solution? It doesn't make netpoll any more robust, but I think in the interests of efficiency it would be fair to require that, when netpolled, a driver must receive frames on the same net device for which it was polled. With this patch we detect that condition and handle it accordingly in e1000_intr. This eliminates the need for us to call the clean_rx method from the poll_controller when napi is configured, instead allowing the poll method to be called from napi_poll, as the netpoll model currently does. This fixes the netdump regression, and eliminates the layering violation and the potential race that we've been discussing. I've just tested it with netdump here and it works quite well. Thoughts appreciated. This looks pretty reasonable, mostly from the perspective that it doesn't put any further ugliness in netpoll. We might want to add a comment somewhere in netpoll of the new rule we're now observing. I'll let the e1000 guys comment on the particulars of the driver change. Signed-off-by: Neil Horman [EMAIL PROTECTED] Signed-off-by: Matt Mackall [EMAIL PROTECTED] -- Mathematics is the supreme nostalgia of our time. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 4/4] ehea: Kconfig and Makefile
Signed-off-by: Jan-Bernd Themann [EMAIL PROTECTED] drivers/net/ehea/Kconfig |6 ++ drivers/net/ehea/Makefile | 45 + 2 files changed, 51 insertions(+) --- linux-2.6.16-rc5-orig/drivers/net/ehea/Makefile1969-12-31 16:00:00.0 -0800 +++ kernel/drivers/net/ehea/Makefile2006-06-07 07:01:14.634178704 -0700 @@ -0,0 +1,45 @@ +# Server eHEA ethernet device driver for Linux on POWER +# +# +# linux/drivers/net/ehea/Makefile +# +# eHEA ethernet device driver for IBM eServer System p +# +# (C) Copyright IBM Corp. 2006 +# +# Authors: +# Christoph Raisch [EMAIL PROTECTED] +# Jan-Bernd Themann [EMAIL PROTECTED] +# Heiko-Joerg Schick [EMAIL PROTECTED] +# Thomas Klein [EMAIL PROTECTED] +# +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + + + + +# To compile module type: +# make -C /root/linux-2.6.5-7.244 SUBDIRS=`pwd` V=1 modules CONFIG_EHEA=m +# make -C /usr/src/linux-2.6.17-rc5 SUBDIRS=`pwd` V=1 modules CONFIG_EHEA=m + +ehea_mod-objs = ehea_main.o ehea_phyp.o ehea_qmr.o ehea_ethtool.o ehea_phyp.o + +obj-$(CONFIG_EHEA) += ehea_mod.o + +#CFLAGS += -Werror + + + --- linux-2.6.16-rc5-orig/drivers/net/ehea/Kconfig1969-12-31 16:00:00.0 -0800 +++ kernel/drivers/net/ehea/Kconfig2006-06-07 07:01:14.637178248 -0700 @@ -0,0 +1,6 @@ +config EHEA + tristate eHEA Ethernet support + depends on IBMEBUS + ---help--- + This driver supports the IBM pSeries + Ethernet adapter - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 3/4] ehea: queue managment
Signed-off-by: Jan-Bernd Themann [EMAIL PROTECTED]
drivers/net/ehea/ehea_qmr.c | 719
drivers/net/ehea/ehea_qmr.h | 390 +++
2 files changed, 1109 insertions(+)
--- linux-2.6.16-rc5-orig/drivers/net/ehea/ehea_qmr.c1969-12-31
16:00:00.0 -0800
+++ kernel/drivers/net/ehea/ehea_qmr.c2006-06-07 07:01:14.664174144 -0700
@@ -0,0 +1,719 @@
+/*
+ * linux/drivers/net/ehea/ehea_qmr.c
+ *
+ * eHEA ethernet device driver for IBM eServer System p
+ *
+ * (C) Copyright IBM Corp. 2006
+ *
+ * Authors:
+ * Christoph Raisch [EMAIL PROTECTED]
+ * Jan-Bernd Themann [EMAIL PROTECTED]
+ * Heiko-Joerg Schick [EMAIL PROTECTED]
+ * Thomas Klein [EMAIL PROTECTED]
+ *
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include ehea.h
+#include ehea_phyp.h
+#include ehea_qmr.h
+
+static void *ipz_qpageit_get_inc(struct ipz_queue *queue)
+{
+void *retvalue = ipz_qeit_get(queue);
+queue-current_q_offset += queue-pagesize;
+if (queue-current_q_offset queue-queue_length) {
+queue-current_q_offset -= queue-pagesize;
+retvalue = NULL;
+}
+else if u64) retvalue) (EHEA_PAGESIZE-1)) != 0) {
+EDEB(4, ERROR!! not at PAGE-Boundary);
+return NULL;
+}
+EDEB(7, queue=%p retvalue=%p, queue, retvalue);
+return retvalue;
+}
+
+static int ipz_queue_ctor(struct ipz_queue *queue,
+ const u32 nr_of_pages,
+ const u32 pagesize, const u32 qe_size,
+ const u32 nr_of_sg)
+{
+int f;
+EDEB_EN(7, nr_of_pages=%x pagesize=%x qe_size=%x,
+nr_of_pages, pagesize, qe_size);
+queue-queue_length = nr_of_pages * pagesize;
+queue-queue_pages = vmalloc(nr_of_pages * sizeof(void *));
+if (!queue-queue_pages) {
+EDEB(4, ERROR!! didn't get the memory);
+return 0;
+}
+memset(queue-queue_pages, 0, nr_of_pages * sizeof(void *));
+
+for (f = 0; f nr_of_pages; f++) {
+(queue-queue_pages)[f] =
+(struct ipz_page *)get_zeroed_page(GFP_KERNEL);
+if (!(queue-queue_pages)[f]) {
+break;
+}
+}
+if (f nr_of_pages) {
+int g;
+EDEB_ERR(4, couldn't get 0ed pages queue=%p f=%x
+ nr_of_pages=%x, queue, f, nr_of_pages);
+for (g = 0; g f; g++) {
+free_page((unsigned long)(queue-queue_pages)[g]);
+}
+return 0;
+}
+queue-current_q_offset = 0;
+queue-qe_size = qe_size;
+queue-act_nr_of_sg = nr_of_sg;
+queue-pagesize = pagesize;
+queue-toggle_state = 1;
+EDEB_EX(7, queue_length=%x queue_pages=%p qe_size=%x
+ act_nr_of_sg=%x, queue-queue_length, queue-queue_pages,
+queue-qe_size, queue-act_nr_of_sg);
+return 1;
+}
+
+static int ipz_queue_dtor(struct ipz_queue *queue)
+{
+int g;
+EDEB_EN(7, ipz_queue pointer=%p, queue);
+if (!queue) {
+return 0;
+}
+if (!queue-queue_pages) {
+return 0;
+}
+EDEB(7, destructing a queue with the following properties:\n
+ queue_length=%x act_nr_of_sg=%x pagesize=%x qe_size=%x,
+ queue-queue_length, queue-act_nr_of_sg, queue-pagesize,
+ queue-qe_size);
+for (g = 0; g (queue-queue_length / queue-pagesize); g++) {
+free_page((unsigned long)(queue-queue_pages)[g]);
+}
+vfree(queue-queue_pages);
+
+EDEB_EX(7, queue freed!);
+return 1;
+}
+
+struct ehea_cq *ehea_cq_new(void)
+{
+struct ehea_cq *cq = vmalloc(sizeof(*cq));
+if (cq)
+memset(cq, 0, sizeof(*cq));
+return cq;
+}
+
+void ehea_cq_delete(struct ehea_cq *cq)
+{
+vfree(cq);
+}
+
+struct ehea_cq *ehea_create_cq(struct ehea_adapter *adapter,
+ int nr_of_cqe, u64 eq_handle, u32 cq_token)
+{
+struct ehea_cq *cq = NULL;
+struct h_galpa gal;
+
+u64 *cq_handle_ref;
+u32 act_nr_of_entries;
+u32 act_pages;
+u64 hret;
+int ipz_rc;
+u32 counter;
+void *vpage = NULL;
+u64 rpage = 0;
+
+EDEB_EN(7, adapter=%p nr_of_cqe=%x , eq_handle: %016lX,
+adapter, nr_of_cqe, eq_handle);
+
+cq = ehea_cq_new();
+if (!cq) {
+cq = NULL;
+EDEB_ERR(4, ehea_create_cq ret=%p (-ENOMEM), cq);
+goto create_cq_exit0;
+}
+
+
Re: Problem authenticating using WPA with bcm43xx-softmac
Johannes Berg wrote:
On Wed, 2006-06-07 at 09:01 -0700, Sam Leffler wrote:
Beware of the order of IE's in the management frames; some AP's are
touchy about this.
Uh oh. I have no idea where the ieee80211 layer sticks that one,
probably at the end.
I moved the WPA IE from the end forward without any effect. I have dumped that IE and find the
following:
SoftMAC: Added WPA_IE of 24 bytes to association request
Contents of WPA_IE: 0xdd 0x16 0x00 0x50 0xf2 0x01 0x01 0x00 0x00 0x50
0xf2 0x02 0x01 0x00 0x00 0x50 0xf2 0x02 0x01 0x00
0x00 0x50 0xf2 0x02
I think I got the right thing, but I'm having trouble interpreting it. The dump was made by
modifying the part of ieee80211softmac_assoc_req in ieee80211softmac_io.c where the WPA IE is added:
/* Add WPA IE */
if (mac-wpa.IElen mac-wpa.IE) {
int i;
memcpy(data, mac-wpa.IE, mac-wpa.IElen);
data += mac-wpa.IElen;
printk(KERN_INFO PFX Added WPA_IE of %d bytes to association
request\n
Contents of WPA_IE: , mac-wpa.IElen);
for (i=0; imac-wpa.IElen; i++)
printk(0x%02x ,mac-wpa.IE[i]);
printk(\n);
}
Larry
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 0/4] ehea: IBM eHEA Ethernet Device Driver
On Wed, 07 Jun 2006 17:12:45 +0200 Jan-Bernd Themann [EMAIL PROTECTED] wrote: Hello, this is the first version of the IBM eHEA Ethernet Device Driver. It supports a new IBM ethernet chip on System p. Main functionality including broadcast multicast and some parts of ethtool already work. Things we're still working on: - vlan support - performance improvements in SMP systems - kernel coding style Please: - Get rid of the debug macro's with 9 arguments... - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
On Wed, 2006-06-07 at 17:57 +0200, Johannes Berg wrote:
On Wed, 2006-06-07 at 17:51 +0200, Johannes Berg wrote:
Well, it should be shown in the 802.11i spec too.
I suppose that it is the association request, and needs to contain the
RSN described in 7.3.2.25 as per 7.2.3.4 in 802.11i. This is, afaik, the
'generic IE' that is added with the wext. Now, it looks like the RSN
isn't included but the WPA2 info or something? Also, the genIE in your
log doesn't look correct to me, starting with ff?? Jouni, do you
have any idea what might be going on?
I believe that wpa_supplicant tells the driver what genie to use through
the SIOCSIWGENIE wext call. The IEs match between what the driver
appears to be reporting, and what wpa_supplicant says from the logs.
wpa_supplicant is almost certainly writing the correct IE to the driver
through wext, so I think the debug output from softmac must be
formatting the string incorrectly when printing it out to the logs.
Looking at it further:
struct ieee80211softmac_wpa {
char *IE;
int IElen;
int IEbuflen;
};
from ieee80211softmac_wx.c: ieee80211softmac_wx_set_genie()
memcpy(mac-wpa.IE, extra, wrqu-data.length);
dprintk(KERN_INFO PFX generic IE set to );
for (i=0;iwrqu-data.length;i++)
dprintk(%.2x, mac-wpa.IE[i]);
dprintk(\n);
the dprintk code isn't doing the right thing here, given an array of
bytes. You probably want:
dprintk(%.2hhx, mac-wpa.IE[i]);
(ie, add the hh before the x to tell the print that it's a char)
Dan
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 3/4] ehea: queue managment
On Wed, 07 Jun 2006 19:05:02 +0200 Jan-Bernd Themann [EMAIL PROTECTED] wrote: Signed-off-by: Jan-Bernd Themann [EMAIL PROTECTED] drivers/net/ehea/ehea_qmr.c | 719 drivers/net/ehea/ehea_qmr.h | 390 +++ 2 files changed, 1109 insertions(+) Make it look less like C++ Do you really need to use vmalloc/vfree? Are the data structures that big? Be careful about global namespace issues. Stick to one prefix like ehea_ for all non static function names. Consider putting all in one file, or use #include to cause it to be one compilation unit. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 4/4] ehea: Kconfig and Makefile
On Wed, 07 Jun 2006 19:06:03 +0200 Jan-Bernd Themann wrote: Signed-off-by: Jan-Bernd Themann [EMAIL PROTECTED] drivers/net/ehea/Kconfig |6 ++ drivers/net/ehea/Makefile | 45 + 2 files changed, 51 insertions(+) --- linux-2.6.16-rc5-orig/drivers/net/ehea/Makefile1969-12-31 16:00:00.0 -0800 +++ kernel/drivers/net/ehea/Makefile2006-06-07 07:01:14.634178704 -0700 @@ -0,0 +1,45 @@ +# Server eHEA ethernet device driver for Linux on POWER +# +# +# linux/drivers/net/ehea/Makefile +# +# eHEA ethernet device driver for IBM eServer System p +# +# (C) Copyright IBM Corp. 2006 +# +# Authors: +# Christoph Raisch [EMAIL PROTECTED] +# Jan-Bernd Themann [EMAIL PROTECTED] +# Heiko-Joerg Schick [EMAIL PROTECTED] +# Thomas Klein [EMAIL PROTECTED] +# +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + + + + +# To compile module type: +# make -C /root/linux-2.6.5-7.244 SUBDIRS=`pwd` V=1 modules CONFIG_EHEA=m +# make -C /usr/src/linux-2.6.17-rc5 SUBDIRS=`pwd` V=1 modules CONFIG_EHEA=m + +ehea_mod-objs = ehea_main.o ehea_phyp.o ehea_qmr.o ehea_ethtool.o ehea_phyp.o + +obj-$(CONFIG_EHEA) += ehea_mod.o + +#CFLAGS += -Werror + + + 45 lines for a 5-line makefile. :( You need to integrate the Makefile and Kconfig into the upper-level directory Makefile and Kconfig files. Don't expect/require people to enter thing like this: +# make -C /root/linux-2.6.5-7.244 SUBDIRS=`pwd` V=1 modules CONFIG_EHEA=m +# make -C /usr/src/linux-2.6.17-rc5 SUBDIRS=`pwd` V=1 modules CONFIG_EHEA=m --- linux-2.6.16-rc5-orig/drivers/net/ehea/Kconfig1969-12-31 16:00:00.0 -0800 +++ kernel/drivers/net/ehea/Kconfig2006-06-07 07:01:14.637178248 -0700 @@ -0,0 +1,6 @@ +config EHEA + tristate eHEA Ethernet support + depends on IBMEBUS + ---help--- + This driver supports the IBM pSeries + Ethernet adapter That last sentence can be on one line, please. --- ~Randy - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
Dan Williams wrote:
On Wed, 2006-06-07 at 17:57 +0200, Johannes Berg wrote:
On Wed, 2006-06-07 at 17:51 +0200, Johannes Berg wrote:
Well, it should be shown in the 802.11i spec too.
I suppose that it is the association request, and needs to contain the
RSN described in 7.3.2.25 as per 7.2.3.4 in 802.11i. This is, afaik, the
'generic IE' that is added with the wext. Now, it looks like the RSN
isn't included but the WPA2 info or something? Also, the genIE in your
log doesn't look correct to me, starting with ff?? Jouni, do you
have any idea what might be going on?
I believe that wpa_supplicant tells the driver what genie to use through
the SIOCSIWGENIE wext call. The IEs match between what the driver
appears to be reporting, and what wpa_supplicant says from the logs.
wpa_supplicant is almost certainly writing the correct IE to the driver
through wext, so I think the debug output from softmac must be
formatting the string incorrectly when printing it out to the logs.
Looking at it further:
struct ieee80211softmac_wpa {
char *IE;
int IElen;
int IEbuflen;
};
from ieee80211softmac_wx.c: ieee80211softmac_wx_set_genie()
memcpy(mac-wpa.IE, extra, wrqu-data.length);
dprintk(KERN_INFO PFX generic IE set to );
for (i=0;iwrqu-data.length;i++)
dprintk(%.2x, mac-wpa.IE[i]);
dprintk(\n);
the dprintk code isn't doing the right thing here, given an array of
bytes. You probably want:
dprintk(%.2hhx, mac-wpa.IE[i]);
(ie, add the hh before the x to tell the print that it's a char)
That doesn't work - the result is
%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx
I changed the line to cast the output byte as a u8 as follows:
dprintk(%.2x, (u8)mac-wpa.IE[i]);
This produces the line
generic IE set to dd160050f2010150f2020150f2020150f202
This is the WPA IE supplied by wpa_supplicant and it matches the one used in the ndiswrapper case.
One mystery solved, but why doesn't it work?
Johannes - should I submit the patch to fix this printout, or would you like to
do it?
Larry
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: 2.6.17-rc6-mm1
On Wed, Jun 07, 2006 at 10:47:24AM -0700, Andrew Morton wrote: ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.17-rc6/2.6.17-rc6-mm1/ Just spotted this typo in a new option. Signed-off-by: Dave Jones [EMAIL PROTECTED] --- rc6mm1/net/Kconfig~ 2006-06-07 14:19:44.0 -0400 +++ rc6mm1/net/Kconfig 2006-06-07 14:19:58.0 -0400 @@ -227,7 +227,7 @@ config NET_TCPPROBE depends on INET EXPERIMENTAL PROC_FS KPROBES ---help--- This module allows for capturing the changes to TCP connection - state in response to incoming patckets. It is used for debugging + state in response to incoming packets. It is used for debugging TCP congestion avoidance modules. If you don't understand what was just said, you don't need it: say N. -- http://www.codemonkey.org.uk - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] e1000: fix netpoll with NAPI
Matt Mackall wrote:
On Wed, Jun 07, 2006 at 11:05:22AM -0400, Neil Horman wrote:
Matt, any ideas on this?
Not at the moment.
how about this for a solution? It doesn't make netpoll any more robust, but I
think in the interests of efficiency it would be fair to require that, when
netpolled, a driver must receive frames on the same net device for which it was
polled. With this patch we detect that condition and handle it accordingly in
e1000_intr. This eliminates the need for us to call the clean_rx method from
the poll_controller when napi is configured, instead allowing the poll method to
be called from napi_poll, as the netpoll model currently does. This fixes the
netdump regression, and eliminates the layering violation and the potential race
that we've been discussing. I've just tested it with netdump here and it works
quite well.
Thoughts appreciated.
This looks pretty reasonable, mostly from the perspective that it
doesn't put any further ugliness in netpoll. We might want to add a
comment somewhere in netpoll of the new rule we're now observing.
I'll let the e1000 guys comment on the particulars of the driver change.
Hi,
we're not too happy with this as it puts a branch right in the regular receive
path. We haven't ran the numbers on it yet but it is likely that this will
lower performance significantly during normal receives for something that is
not common use.
Attached below a (revised) patch that adds proper locking around the rx_clean
to prevent the race.
Cheers,
Auke
---
e1000: fix netpoll with NAPI
This fixes netpoll when using NAPI, and protects against a rare race condition
in the netpoll routine, while staying out of our ways from the normal hotpath.
Signed-off-by: Mitch Williams [EMAIL PROTECTED]
Signed-off-by: Auke Kok [EMAIL PROTECTED]
---
e1000_main.c | 17 -
1 file changed, 16 insertions(+), 1 deletion(-)
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index bd709e5..876251c 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -4584,10 +4584,25 @@ static void
e1000_netpoll(struct net_device *netdev)
{
struct e1000_adapter *adapter = netdev_priv(netdev);
+#ifdef CONFIG_E1000_NAPI
+ int budget = 0;
+
+ disable_irq(adapter-pdev-irq);
+ if (likely(netif_rx_schedule_prep(adapter-polling_netdev[0]))) {
+ if (spin_trylock(adapter-tx_queue_lock)) {
+ e1000_clean_tx_irq(adapter, adapter-tx_ring[0]);
+ spin_unlock(adapter-tx_queue_lock);
+ }
+ adapter-clean_rx(adapter, adapter-rx_ring,
+budget, netdev-weight);
+ clear_bit(__LINK_STATE_RX_SCHED,
+adapter-polling_netdev[0].state);
+ }
+#else
+
disable_irq(adapter-pdev-irq);
e1000_intr(adapter-pdev-irq, netdev, NULL);
e1000_clean_tx_irq(adapter, adapter-tx_ring);
-#ifndef CONFIG_E1000_NAPI
adapter-clean_rx(adapter, adapter-rx_ring);
#endif
enable_irq(adapter-pdev-irq);
Re: [PATCH 1/2] e1000: fix netpoll with NAPI
== Regarding Re: [PATCH 1/2] e1000: fix netpoll with NAPI; Auke Kok [EMAIL
PROTECTED] adds:
auke-jan.h.kok Hi,
auke-jan.h.kok we're not too happy with this as it puts a branch right in
auke-jan.h.kok the regular receive path. We haven't ran the numbers on it
auke-jan.h.kok yet but it is likely that this will lower performance
auke-jan.h.kok significantly during normal receives for something that is
auke-jan.h.kok not common use.
auke-jan.h.kok Attached below a (revised) patch that adds proper locking
auke-jan.h.kok around the rx_clean to prevent the race.
That patch locks around the tx clean routine. As such, it doesn't prevent
the problem.
+ disable_irq(adapter-pdev-irq);
+ if (likely(netif_rx_schedule_prep(adapter-polling_netdev[0]))) {
+ if (spin_trylock(adapter-tx_queue_lock)) {
+ e1000_clean_tx_irq(adapter, adapter-tx_ring[0]);
+ spin_unlock(adapter-tx_queue_lock);
+ }
+ adapter-clean_rx(adapter, adapter-rx_ring,
+ budget, netdev-weight);
+ clear_bit(__LINK_STATE_RX_SCHED,
+ adapter-polling_netdev[0].state);
-Jeff
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] e1000: fix netpoll with NAPI
On Wed, Jun 07, 2006 at 11:25:29AM -0700, Auke Kok wrote:
@@ -4584,10 +4584,25 @@ static void
e1000_netpoll(struct net_device *netdev)
{
struct e1000_adapter *adapter = netdev_priv(netdev);
+#ifdef CONFIG_E1000_NAPI
+ int budget = 0;
+
+ disable_irq(adapter-pdev-irq);
+ if (likely(netif_rx_schedule_prep(adapter-polling_netdev[0]))) {
+ if (spin_trylock(adapter-tx_queue_lock)) {
+ e1000_clean_tx_irq(adapter, adapter-tx_ring[0]);
+ spin_unlock(adapter-tx_queue_lock);
+ }
+ adapter-clean_rx(adapter, adapter-rx_ring,
+ budget, netdev-weight);
+ clear_bit(__LINK_STATE_RX_SCHED,
+ adapter-polling_netdev[0].state);
+ }
+#else
+
disable_irq(adapter-pdev-irq);
Pedantic objection, but I think this would read easier w/o the extra
newline before disable_irq.
John
--
John W. Linville
[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] e1000: fix netpoll with NAPI
On Wed, Jun 07, 2006 at 02:44:54PM -0400, Jeff Moyer wrote:
== Regarding Re: [PATCH 1/2] e1000: fix netpoll with NAPI; Auke Kok [EMAIL
PROTECTED] adds:
auke-jan.h.kok Hi,
auke-jan.h.kok we're not too happy with this as it puts a branch right in
auke-jan.h.kok the regular receive path. We haven't ran the numbers on it
auke-jan.h.kok yet but it is likely that this will lower performance
auke-jan.h.kok significantly during normal receives for something that is
auke-jan.h.kok not common use.
auke-jan.h.kok Attached below a (revised) patch that adds proper locking
auke-jan.h.kok around the rx_clean to prevent the race.
That patch locks around the tx clean routine. As such, it doesn't prevent
the problem.
Further to that, do tests on this if you like, but I would certainly think a
properly formed conditional operation is going to provide better performance
than a spin_lock operation in the receive path. Why not just turn the:
if(netpoll_op)
into an
if(unlikely(netpoll_op))
I expect that will reduce the overhead of the conditional to effectively zero
for the normal receive case. The following patch does that, and I expect you
performance won't suffer at all:
Signed-off-by: Neil Horman [EMAIL PROTECTED]
e1000_main.c | 54 +-
1 files changed, 45 insertions(+), 9 deletions(-)
--- linux-2.6.9/drivers/net/e1000/e1000_main.c.neil 2006-06-06
10:37:42.0 -0400
+++ linux-2.6.9/drivers/net/e1000/e1000_main.c 2006-06-07 10:48:22.0
-0400
@@ -3207,8 +3207,9 @@ e1000_update_stats(struct e1000_adapter
* @pt_regs: CPU registers structure
**/
+
static irqreturn_t
-e1000_intr(int irq, void *data, struct pt_regs *regs)
+__e1000_intr(int irq, void *data, struct pt_regs *regs, int netpoll_op)
{
struct net_device *netdev = data;
struct e1000_adapter *adapter = netdev_priv(netdev);
@@ -3217,6 +3218,7 @@ e1000_intr(int irq, void *data, struct p
#ifndef CONFIG_E1000_NAPI
int i;
#else
+ struct net_device *dev_to_sched;
/* Interrupt Auto-Mask...upon reading ICR,
* interrupts are masked. No need for the
* IMC write, but it does mean we should
@@ -3255,8 +3257,22 @@ e1000_intr(int irq, void *data, struct p
E1000_WRITE_REG(hw, IMC, ~0);
E1000_WRITE_FLUSH(hw);
}
- if (likely(netif_rx_schedule_prep(adapter-polling_netdev[0])))
- __netif_rx_schedule(adapter-polling_netdev[0]);
+
+ /*
+* netpoll operations, in the interests of efficiency
+* only do napi polling on the device passed to the
+* poll_controller. Therefore, if we are preforming
+* a netpoll operation, then we can't schedule a receive
+* to one of the dummy net devices that exist for sole
+* purpose of spreading out rx schedules
+*/
+ if (unlikely(netpoll_op))
+ dev_to_sched = netdev;
+ else
+ dev_to_sched = adapter-polling_netdev[0];
+
+ if (likely(netif_rx_schedule_prep(dev_to_sched)))
+ __netif_rx_schedule(dev_to_sched);
else
e1000_irq_enable(adapter);
#else
@@ -3288,6 +3304,13 @@ e1000_intr(int irq, void *data, struct p
return IRQ_HANDLED;
}
+static irqreturn_t
+e1000_intr(int irq, void *data, struct pt_regs *regs)
+{
+ return __e1000_intr(irq, data, regs, 0);
+}
+
+
#ifdef CONFIG_E1000_NAPI
/**
* e1000_clean - NAPI Rx polling callback
@@ -3300,7 +3323,6 @@ e1000_clean(struct net_device *poll_dev,
struct e1000_adapter *adapter;
int work_to_do = min(*budget, poll_dev-quota);
int tx_cleaned = 0, i = 0, work_done = 0;
-
/* Must NOT use netdev_priv macro here. */
adapter = poll_dev-priv;
@@ -3308,10 +3330,24 @@ e1000_clean(struct net_device *poll_dev,
if (!netif_carrier_ok(adapter-netdev))
goto quit_polling;
- while (poll_dev != adapter-polling_netdev[i]) {
- i++;
- if (unlikely(i == adapter-num_rx_queues))
- BUG();
+ /*
+* only search for a matching polling_netdev in the event
+* that this isn't a real registered net_device
+* A real net device can be passed in here in the event
+* that netdump has been activated (this comes through
+* netpoll_poll_dev). We detect this by virtue of the
+* fact that each polling_netdev-priv points to the private
+* data of its parent (registered) netdev. So if:
+* poll_dev-priv == netdev_priv(poll_dev), its a real device
+* otherwise its a polling_netdev.
+*/
+ if (likely(adapter != netdev_priv(poll_dev))) {
+ while (poll_dev != adapter-polling_netdev[i]) {
+ i++;
+ if (unlikely(i == adapter-num_rx_queues))
+ BUG();
+ }
+
RE: kernel panic (on DHCP discover?) in sky2 driver of 2.6.17-rc1
I was perhaps a bit quick to declare victory. While the results below stand and
the machine survived the last few days (idle), it occurred to me only today, to
have a look at the kernel's message buffer, where I found following:
--8--
sky2 eth0: enabling interface
sky2 eth0: Link is up at 1000 Mbps, full duplex, flow control none
sky2 eth1: enabling interface
sky2 eth1: Link is up at 1000 Mbps, full duplex, flow control none
audit(1149379670.514:3): audit_pid=1915 old=0 by auid=4294967295
unknown: hw csum failure.
sky2 eth1: rx error, status 0x7ffc0001 length 444
Call Trace: 811de741{__skb_checksum_complete+76}
812030cb{__tcp_checksum_complete_user+33}
812080d8{tcp_rcv_established+817} 8120f3ee{tcp_v4_
do_rcv+43}
811da2ee{sk_wait_data+203} 811fe5a8{tcp_prequeue_p
rocess+121}
811ff71d{tcp_recvmsg+1104} 811d9712{sock_common_re
cvmsg+48}
811d7d4f{do_sock_read+209} 811d7e7e{sock_aio_read+
83}
811e2ca1{dev_queue_xmit+0} 8106dce9{do_sync_read+1
99}
8103d699{remove_wait_queue+18} 8103d530{autoremove
_wake_function+0}
8106df83{vfs_read+228} 8106ea12{sys_read+69}
81009b0d{tracesys+209}
unknown: hw csum failure.
sky2 eth1: rx error, status 0x7ffc0001 length 444
Call Trace: 811de741{__skb_checksum_complete+76}
812030cb{__tcp_checksum_complete_user+33}
812080d8{tcp_rcv_established+817} 8120f3ee{tcp_v4_
do_rcv+43}
811da2ee{sk_wait_data+203} 811fe5a8{tcp_prequeue_p
rocess+121}
811ff71d{tcp_recvmsg+1104} 811d9712{sock_common_re
cvmsg+48}
811d7696{alloc_sock_iocb+20} 811d7d4f{do_sock_read
+209}
811d7e7e{sock_aio_read+83} 8106dce9{do_sync_read+1
99}
8103d699{remove_wait_queue+18} 8103d530{autoremove
_wake_function+0}
8106df83{vfs_read+228} 8106ea12{sys_read+69}
81009b0d{tracesys+209}
unknown: hw csum failure.
sky2 eth1: rx error, status 0x7ffc0001 length 444
Call Trace: 811de741{__skb_checksum_complete+76}
812030cb{__tcp_checksum_complete_user+33}
812080d8{tcp_rcv_established+817} 8120f3ee{tcp_v4_
do_rcv+43}
811da2ee{sk_wait_data+203} 811fe5a8{tcp_prequeue_p
rocess+121}
811ff71d{tcp_recvmsg+1104} 811d9712{sock_common_re
cvmsg+48}
811d7d4f{do_sock_read+209} 811d7e7e{sock_aio_read+
83}
8106dce9{do_sync_read+199} 8103d699{remove_wait_qu
eue+18}
8103d530{autoremove_wake_function+0} 8102f1a0{curr
ent_kernel_time+13}
8106df83{vfs_read+228} 8106ea12{sys_read+69}
81009b0d{tracesys+209}
sky2 eth0: rx error, status 0x7ffc0001 length 444
sky2 eth0: rx error, status 0x7ffc0001 length 444
sky2 eth1: rx error, status 0x7ffc0001 length 444
sky2 eth1: rx error, status 0x7ffc0001 length 444
sky2 eth1: rx error, status 0x7ffc0001 length 444
--8--
Looks, like we're almost, but not quite there yet.
cheers
Guenther
-Original Message-
From: Guenther Thomsen
Sent: Saturday, June 03, 2006 9:06 PM
To: 'Stephen Hemminger'
Cc: John W. Linville; netdev@vger.kernel.org
Subject: RE: kernel panic (on DHCP discover?) in sky2 driver of
2.6.17-rc1
I received the hardware back and took the opportunity to test with
2.6.17-rc5-git11. So far I did only little tests (ttcp on both interfaces in,
out or mixed with some 10e6 packets), but it looks good. No errors (well, 16
overruns in 76574513 packets) and line rate (about 111MB/s) on both channels
simultaneously. Hurray!
Thanks a lot for you continued efforts.
Guenther
-Original Message-
From: Stephen Hemminger [mailto:[EMAIL PROTECTED]
Sent: Tuesday, May 16, 2006 12:12 PM
To: Guenther Thomsen
Cc: John W. Linville; netdev@vger.kernel.org
Subject: Re: kernel panic (on DHCP discover?) in sky2 driver of
2.6.17-rc1
Could you try the 2.6.17-rc4 version with this patch. It turns out the board
seems to give out of order status responses.
Ignore the vendor sk98lin driver, when I try the stock version it spends it's
life resetting itself because it sets up PCI bus wrong. If I fix that, it spends
it's time getting confused because it can't handle intermixed status reports
properly (checksum et all is per port not per board).
drivers/net/sky2.c | 28 +---
1 files changed, 21 insertions(+), 7 deletions(-)
792547bc5e8e4f7d5a1070a168056f429635c254
diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index ffd267f..11e7914 100644
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -1020,8 +1020,27 @@ static int sky2_up(struct net_device *de
struct sky2_hw *hw = sky2-hw;
unsigned port = sky2-port;
u32 ramsize, rxspace, imask;
- int
Re: Problem authenticating using WPA with bcm43xx-softmac
On Wed, 2006-06-07 at 13:12 -0500, Larry Finger wrote: Dan Williams wrote: from ieee80211softmac_wx.c: ieee80211softmac_wx_set_genie() memcpy(mac-wpa.IE, extra, wrqu-data.length); dprintk(KERN_INFO PFX generic IE set to ); for (i=0;iwrqu-data.length;i++) dprintk(%.2x, mac-wpa.IE[i]); dprintk(\n); the dprintk code isn't doing the right thing here, given an array of bytes. You probably want: dprintk(%.2hhx, mac-wpa.IE[i]); (ie, add the hh before the x to tell the print that it's a char) That doesn't work - the result is Weird, does the kernel not do something that fprintf() _does_ do here? I tested with a short C program that mimics the behavior of this chunk of code, and %.2x didn't work, but %.2hhx certainly did. hh is supposed to mean A following integer conversion corresponds to a signed char or unsigned char argument. The original conversion was converting stuff from la-la land after the first 4 bytes (in both softmac and my testcase), and hh solved it in the testcase. I did try casting to char, but glibc pretty much ignored that. Dan %hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx%hx I changed the line to cast the output byte as a u8 as follows: dprintk(%.2x, (u8)mac-wpa.IE[i]); This produces the line generic IE set to dd160050f2010150f2020150f2020150f202 This is the WPA IE supplied by wpa_supplicant and it matches the one used in the ndiswrapper case. One mystery solved, but why doesn't it work? Johannes - should I submit the patch to fix this printout, or would you like to do it? Larry - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Updated sysctl documentation take #2
Since people didn't like the many small files approach, I've moved it to directories containing index.txt files: Documentation/sysctl/vm/index.txt Documentation/sysctl/net/core/index.txt Documentation/sysctl/net/unix/index.txt Documentation/sysctl/net/ipv4/index.txt Documentation/sysctl/net/ipv4/conf/index.txt Documentation/sysctl/net/ipv4/route/index.txt Documentation/sysctl/net/ipv4/neigh/index.txt and so on. As previously, this moves all sysctl documentation (including XFS and network) to Documentation/sysctl/*. The patch is against linus tree and weights more than 200K in size and is place at: http://www.terra.es/personal/diegocg/sysctl-docs Comments? - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] pcnet32 driver NAPI support
On Wed, Jun 07, 2006 at 11:20:40AM -0700, Don Fry wrote: I am also working on a NAPI version of the pcnet32 driver for many of the same reasons, and will compare what you have with my own implementation. I probably won't be able to do much until Friday. Just a couple of comments. I am adding netdev@vger.kernel.org to the cc list, as most network driver discussion is done here rather than lkml. linux-kernel (and linux-net) should be deleted in future replies. I must have picked the wrong place to cc. The 2.6.17-rc6 would be the correct source to patch against. Since this is an enhancement it will not come out till 2.6.18. I thought so. That is why I did it against both 2.6.17-rc6 and 2.6.16 (since I use it with 2.6.16). I would not change the driver name from pcnet32 to pcnet32napi, but I would changes the version from 1.32 to 1.33NAPI or something like that. Hmm, perhaps. I just wanted something that made it obvious in dmesg which driver I was running. I see tulip actually does put it in the version instead. I don't remember where I got the driver name change idea from. Some areas of concern that you may have addressed already, I have not scanned your changes yet, are what happens if the ring size is changed without bringing down the interface (via ethtool), or if the loopback test is run in a similar fashion, or a tx timeout occurs. The same thing as if it was done before enabling napi. From a few messages I have seen, it doesn't work right now, and it won't work any better with my changes. I have never tried changing the ring size on the fly, so I don't know. It appears that the port is stopped before the ring size change is done, although I can't really tell how it handles things if the queue is not empty when it stops the port. Does it try to handle anything left in the ring first or does it just toss those packets? (That I would consider wrong). The lp-lock MUST be held whenever accessing the csr or bcr registers as this is a multi-step process, and has been the source of problems in the past. Even on UP systems. Hmm, I just followed what appeared to be in pcnet32_rx and how tulip and a few other drivers had done their napi conversions. It certainly works for me the way I did it. Haven't seen any lockups yet. I do see that I am not holding the lock when I acknowledge IRQs in pcnet32_poll, which pcnet32_rx doesn't need to worry about since it is called from the interrupt handler which already holds the lock. That should be fixed then. So I can do: // Clear RX interrupts spin_lock(lp-lock); lp-a.write_csr (ioaddr, 0, 0x1400); spin_unlock(lp-lock); That part seems simple enough to protect. Is this safe without holding the lock? } while(lp-a.read_csr (ioaddr, 0) 0x1400); Not sure how to wrap a lock around that one without holding the lock for way too long. perhaps: spin_lock(lp-lock); state=lp-a.read_csr (ioaddr, 0) 0x1400; spin_unlock(lp-lock); } while(state); Does that seem more reasonable? Len Sorensen - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: kernel panic (on DHCP discover?) in sky2 driver of 2.6.17-rc1
On Wed, 7 Jun 2006 12:33:21 -0700
Guenther Thomsen [EMAIL PROTECTED] wrote:
I was perhaps a bit quick to declare victory. While the results below stand
and the machine survived the last few days (idle), it occurred to me only
today, to have a look at the kernel's message buffer, where I found following:
--8--
sky2 eth0: enabling interface
sky2 eth0: Link is up at 1000 Mbps, full duplex, flow control none
sky2 eth1: enabling interface
sky2 eth1: Link is up at 1000 Mbps, full duplex, flow control none
audit(1149379670.514:3): audit_pid=1915 old=0 by auid=4294967295
unknown: hw csum failure.
sky2 eth1: rx error, status 0x7ffc0001 length 444
Call Trace: 811de741{__skb_checksum_complete+76}
812030cb{__tcp_checksum_complete_user+33}
812080d8{tcp_rcv_established+817} 8120f3ee{tcp_v4_
do_rcv+43}
811da2ee{sk_wait_data+203} 811fe5a8{tcp_prequeue_p
rocess+121}
811ff71d{tcp_recvmsg+1104} 811d9712{sock_common_re
cvmsg+48}
811d7d4f{do_sock_read+209} 811d7e7e{sock_aio_read+
83}
811e2ca1{dev_queue_xmit+0} 8106dce9{do_sync_read+1
99}
8103d699{remove_wait_queue+18} 8103d530{autoremove
_wake_function+0}
8106df83{vfs_read+228} 8106ea12{sys_read+69}
81009b0d{tracesys+209}
unknown: hw csum failure.
sky2 eth1: rx error, status 0x7ffc0001 length 444
Different problem, I have seen it before. Basically if the receiver gets
overloaded, the
packet FIFO gets full. The driver needs to have some kind of recovery logic for
this;
probably just shutting down the receiver and restarting.
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Problem authenticating using WPA with bcm43xx-softmac
Dan Williams wrote: Weird, does the kernel not do something that fprintf() _does_ do here? I tested with a short C program that mimics the behavior of this chunk of code, and %.2x didn't work, but %.2hhx certainly did. hh is supposed to mean A following integer conversion corresponds to a signed char or unsigned char argument. The original conversion was converting stuff from la-la land after the first 4 bytes (in both softmac and my testcase), and hh solved it in the testcase. I did try casting to char, but glibc pretty much ignored that. I also found that casting with char gave the same result as no cast, but that a u8 cast worked. I guess printk is _almost_ like fprintf. Larry - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Updated sysctl documentation take #2
On Wed, 7 Jun 2006 20:53:16 +0200 Diego Calleja wrote: Since people didn't like the many small files approach, I've moved it to directories containing index.txt files: Documentation/sysctl/vm/index.txt Documentation/sysctl/net/core/index.txt Documentation/sysctl/net/unix/index.txt Documentation/sysctl/net/ipv4/index.txt Documentation/sysctl/net/ipv4/conf/index.txt Documentation/sysctl/net/ipv4/route/index.txt Documentation/sysctl/net/ipv4/neigh/index.txt and so on. As previously, this moves all sysctl documentation (including XFS and network) to Documentation/sysctl/*. The patch is against linus tree and weights more than 200K in size and is place at: http://www.terra.es/personal/diegocg/sysctl-docs ~ diffstat -p1 -w 70 sysctl-docs Documentation/filesystems/proc.txt| 1210 Documentation/filesystems/xfs.txt | 84 - Documentation/networking/00-INDEX |2 Documentation/networking/Configurable |2 Documentation/networking/decnet.txt |4 Documentation/networking/ip-sysctl.txt| 899 --- Documentation/networking/xfrm_sync.txt| 11 Documentation/sysctl/README | 121 - Documentation/sysctl/abi.txt | 54 Documentation/sysctl/abi/index.txt| 49 Documentation/sysctl/dev/README |2 Documentation/sysctl/fs.txt | 150 - Documentation/sysctl/fs/index.txt | 240 +++ Documentation/sysctl/fs/xfs/index.txt | 170 ++ Documentation/sysctl/kernel.txt | 344 Documentation/sysctl/kernel/index.txt | 628 Documentation/sysctl/net/README |8 Documentation/sysctl/net/appletalk/index.txt | 38 Documentation/sysctl/net/bridge/index.txt | 44 Documentation/sysctl/net/core/index.txt | 115 + Documentation/sysctl/net/ipv4/conf/index.txt | 270 +++ Documentation/sysctl/net/ipv4/index.txt | 739 + Documentation/sysctl/net/ipv4/neigh/index.txt | 138 + Documentation/sysctl/net/ipv4/route/index.txt | 143 + Documentation/sysctl/net/ipv6/conf/index.txt | 256 +++ Documentation/sysctl/net/ipv6/icmp/index.txt | 15 Documentation/sysctl/net/ipv6/index.txt | 65 Documentation/sysctl/net/ipv6/neigh/index.txt | 134 + Documentation/sysctl/net/ipv6/route/index.txt | 78 + Documentation/sysctl/net/unix/index.txt | 15 Documentation/sysctl/sunrpc.txt | 20 Documentation/sysctl/sunrpc/index.txt | 64 Documentation/sysctl/vm.txt | 180 -- Documentation/sysctl/vm/index.txt | 334 Documentation/sysrq.txt | 20 35 files changed, 3611 insertions(+), 3035 deletions(-) I don't know how long it takes to review such a large patch, but I'll continue to do so. For now: README: 1. +Documentation for /proc/sys/, aka. sysctl a.k.a. or just aka. Not aka.. or spell it out, or omit it, maybe like: Documentation for /proc/sys/ (sysctl) 2. Limit lines to max. of 80 characters, but around 70-72 is better IMO. That allows someone to make minor corrections without having to fudge the lines around. 3. +This means there're several parameters use there are 4. +are: enabling or disabling forwading in a certain network interface forwarding 5. +is also used to export some stadistic information, ej: some statistic or statistics or statistical ej: ?? use e.g.: (preferred) or eg: 6. +can be read but can _not_ be written. _cannot_ 7. +own program to read them aswell) as well) 8. +to change '/' by '.' and ignore the '/proc/sys' part of the path, ie s/by/to/ s/ie/i.e./ 9. +means that any tweak that you do will be lost the next time you restart your * collapse 2 spaces to 1 10. +_VERY_ DANGEROUS and can ruin the performance performance, drop one performance 11. +As a quick 'ls /proc/sys' will show, the directory consists of several subdirs. +Each subdir is mainly about one part of the kernel, so you can do configuration Spell out subdirectories and subdirectory. 12. +fs/specific filesystems filehandle, inode, dentry and quota tuning insert : after filesystems OK, that's all for the README file. I'll look at the rest of it sometime this week. I don't think that it's quite ready to be merged. --- ~Randy - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH v2 2/2] iWARP Core Changes.
This patch contains modifications to the existing rdma header files,
core files, drivers, and ulp files to support iWARP.
Review updates:
- copy_addr() - rdma_copy_addr()
- dst_dev_addr param in rdma_copy_addr to const.
- various spacing nits with recasting
- include linux/inetdevice.h to get ip_dev_find() prototype.
---
drivers/infiniband/core/Makefile |4
drivers/infiniband/core/addr.c | 19 +
drivers/infiniband/core/cache.c |8 -
drivers/infiniband/core/cm.c |3
drivers/infiniband/core/cma.c| 353 +++---
drivers/infiniband/core/device.c |6
drivers/infiniband/core/mad.c| 11 +
drivers/infiniband/core/sa_query.c |5
drivers/infiniband/core/smi.c| 18 +
drivers/infiniband/core/sysfs.c | 18 +
drivers/infiniband/core/ucm.c|5
drivers/infiniband/core/user_mad.c |9 -
drivers/infiniband/hw/ipath/ipath_verbs.c|2
drivers/infiniband/hw/mthca/mthca_provider.c |2
drivers/infiniband/ulp/ipoib/ipoib_main.c|8 +
drivers/infiniband/ulp/srp/ib_srp.c |2
include/rdma/ib_addr.h | 15 +
include/rdma/ib_verbs.h | 39 +++
18 files changed, 435 insertions(+), 92 deletions(-)
diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile
index 68e73ec..163d991 100644
--- a/drivers/infiniband/core/Makefile
+++ b/drivers/infiniband/core/Makefile
@@ -1,7 +1,7 @@
infiniband-$(CONFIG_INFINIBAND_ADDR_TRANS) := ib_addr.o rdma_cm.o
obj-$(CONFIG_INFINIBAND) +=ib_core.o ib_mad.o ib_sa.o \
- ib_cm.o $(infiniband-y)
+ ib_cm.o iw_cm.o $(infiniband-y)
obj-$(CONFIG_INFINIBAND_USER_MAD) += ib_umad.o
obj-$(CONFIG_INFINIBAND_USER_ACCESS) +=ib_uverbs.o ib_ucm.o
@@ -14,6 +14,8 @@ ib_sa-y :=sa_query.o
ib_cm-y := cm.o
+iw_cm-y := iwcm.o
+
rdma_cm-y := cma.o
ib_addr-y := addr.o
diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c
index d294bbc..83f84ef 100644
--- a/drivers/infiniband/core/addr.c
+++ b/drivers/infiniband/core/addr.c
@@ -32,6 +32,7 @@ #include linux/mutex.h
#include linux/inetdevice.h
#include linux/workqueue.h
#include linux/if_arp.h
+#include linux/inetdevice.h
#include net/arp.h
#include net/neighbour.h
#include net/route.h
@@ -60,12 +61,15 @@ static LIST_HEAD(req_list);
static DECLARE_WORK(work, process_req, NULL);
static struct workqueue_struct *addr_wq;
-static int copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
-unsigned char *dst_dev_addr)
+int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
+const unsigned char *dst_dev_addr)
{
switch (dev-type) {
case ARPHRD_INFINIBAND:
- dev_addr-dev_type = IB_NODE_CA;
+ dev_addr-dev_type = RDMA_NODE_IB_CA;
+ break;
+ case ARPHRD_ETHER:
+ dev_addr-dev_type = RDMA_NODE_RNIC;
break;
default:
return -EADDRNOTAVAIL;
@@ -77,6 +81,7 @@ static int copy_addr(struct rdma_dev_add
memcpy(dev_addr-dst_dev_addr, dst_dev_addr, MAX_ADDR_LEN);
return 0;
}
+EXPORT_SYMBOL(rdma_copy_addr);
int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr)
{
@@ -88,7 +93,7 @@ int rdma_translate_ip(struct sockaddr *a
if (!dev)
return -EADDRNOTAVAIL;
- ret = copy_addr(dev_addr, dev, NULL);
+ ret = rdma_copy_addr(dev_addr, dev, NULL);
dev_put(dev);
return ret;
}
@@ -160,7 +165,7 @@ static int addr_resolve_remote(struct so
/* If the device does ARP internally, return 'done' */
if (rt-idev-dev-flags IFF_NOARP) {
- copy_addr(addr, rt-idev-dev, NULL);
+ rdma_copy_addr(addr, rt-idev-dev, NULL);
goto put;
}
@@ -180,7 +185,7 @@ static int addr_resolve_remote(struct so
src_in-sin_addr.s_addr = rt-rt_src;
}
- ret = copy_addr(addr, neigh-dev, neigh-ha);
+ ret = rdma_copy_addr(addr, neigh-dev, neigh-ha);
release:
neigh_release(neigh);
put:
@@ -244,7 +249,7 @@ static int addr_resolve_local(struct soc
if (ZERONET(src_ip)) {
src_in-sin_family = dst_in-sin_family;
src_in-sin_addr.s_addr = dst_ip;
- ret = copy_addr(addr, dev, dev-dev_addr);
+ ret = rdma_copy_addr(addr, dev, dev-dev_addr);
} else if (LOOPBACK(src_ip)) {
ret = rdma_translate_ip((struct sockaddr *)dst_in, addr);
if (!ret)
diff --git a/drivers/infiniband/core/cache.c
[PATCH v2 0/7][RFC] Ammasso 1100 iWARP Driver
This patchset implements the iWARP provider driver for the Ammasso 1100 RNIC. It is dependent on the iWARP Core Support patch set. We're submitting it for review with the goal for inclusion in the 2.6.19 kernel. This code has gone through several reviews in the openib-general list. Now we are submitting it for external review by the linux community. This StGIT patchset is cloned from Roland Dreier's infiniband.git for-2.6.18 branch. The patchset consists of 7 patches: 1 - Low-level device interface and native stack support 2 - Work request definitions 3 - Provider interface 4 - Memory management 5 - User mode message queue implementation 6 - Verbs queue implementation 7 - Kconfig and Makefile I believe I've addressed all the round 1 review comments. Details of the changes are tracked in each patch comment. Signed-off-by: Tom Tucker [EMAIL PROTECTED] Signed-off-by: Steve Wise [EMAIL PROTECTED] - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH v2 0/2][RFC] iWARP Core Support
This patchset defines the modifications to the Linux infiniband subsystem to support iWARP devices. We're submitting it for review now with the goal for inclusion in the 2.6.19 kernel. This code has gone through several reviews in the openib-general list. Now we are submitting it for external review by the linux community. This StGIT patchset is cloned from Roland Dreier's infiniband.git for-2.6.18 branch. The patchset consists of 2 patches: 1 - New iWARP CM implementation. 2 - Core changes to support iWARP. I believe I've addressed all the round 1 review comments. Details of the changes are tracked in each patch comment. Signed-off-by: Tom Tucker [EMAIL PROTECTED] Signed-off-by: Steve Wise [EMAIL PROTECTED] - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH v2 5/7] AMSO1100 Message Queues.
Review Changes:
- remove useless asserts
- assert() - BUG_ON()
- C2_DEBUG - DEBUG
---
drivers/infiniband/hw/amso1100/c2_mq.c | 175
drivers/infiniband/hw/amso1100/c2_mq.h | 103 +++
2 files changed, 278 insertions(+), 0 deletions(-)
diff --git a/drivers/infiniband/hw/amso1100/c2_mq.c
b/drivers/infiniband/hw/amso1100/c2_mq.c
new file mode 100644
index 000..0b0ab02
--- /dev/null
+++ b/drivers/infiniband/hw/amso1100/c2_mq.c
@@ -0,0 +1,175 @@
+/*
+ * Copyright (c) 2005 Ammasso, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer in the documentation and/or other materials
+ *provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include c2.h
+#include c2_mq.h
+
+void *c2_mq_alloc(struct c2_mq *q)
+{
+ BUG_ON(q-magic != C2_MQ_MAGIC);
+ BUG_ON(q-type != C2_MQ_ADAPTER_TARGET);
+
+ if (c2_mq_full(q)) {
+ return NULL;
+ } else {
+#ifdef DEBUG
+ struct c2wr_hdr *m =
+ (struct c2wr_hdr *) (q-msg_pool.host + q-priv *
q-msg_size);
+#ifdef CCMSGMAGIC
+ BUG_ON(m-magic != be32_to_cpu(~CCWR_MAGIC));
+ m-magic = cpu_to_be32(CCWR_MAGIC);
+#endif
+ return m;
+#else
+ return q-msg_pool.host + q-priv * q-msg_size;
+#endif
+ }
+}
+
+void c2_mq_produce(struct c2_mq *q)
+{
+ BUG_ON(q-magic != C2_MQ_MAGIC);
+ BUG_ON(q-type != C2_MQ_ADAPTER_TARGET);
+
+ if (!c2_mq_full(q)) {
+ q-priv = (q-priv + 1) % q-q_size;
+ q-hint_count++;
+ /* Update peer's offset. */
+ __raw_writew(cpu_to_be16(q-priv), q-peer-shared);
+ }
+}
+
+void *c2_mq_consume(struct c2_mq *q)
+{
+ BUG_ON(q-magic != C2_MQ_MAGIC);
+ BUG_ON(q-type != C2_MQ_HOST_TARGET);
+
+ if (c2_mq_empty(q)) {
+ return NULL;
+ } else {
+#ifdef DEBUG
+ struct c2wr_hdr *m = (struct c2wr_hdr *)
+ (q-msg_pool.host + q-priv * q-msg_size);
+#ifdef CCMSGMAGIC
+ BUG_ON(m-magic != be32_to_cpu(CCWR_MAGIC));
+#endif
+ return m;
+#else
+ return q-msg_pool.host + q-priv * q-msg_size;
+#endif
+ }
+}
+
+void c2_mq_free(struct c2_mq *q)
+{
+ BUG_ON(q-magic != C2_MQ_MAGIC);
+ BUG_ON(q-type != C2_MQ_HOST_TARGET);
+
+ if (!c2_mq_empty(q)) {
+
+#ifdef CCMSGMAGIC
+ {
+ struct c2wr_hdr __iomem *m = (struct c2wr_hdr __iomem *)
+ (q-msg_pool.adapter + q-priv * q-msg_size);
+ __raw_writel(cpu_to_be32(~CCWR_MAGIC), m-magic);
+ }
+#endif
+ q-priv = (q-priv + 1) % q-q_size;
+ /* Update peer's offset. */
+ __raw_writew(cpu_to_be16(q-priv), q-peer-shared);
+ }
+}
+
+
+void c2_mq_lconsume(struct c2_mq *q, u32 wqe_count)
+{
+ BUG_ON(q-magic != C2_MQ_MAGIC);
+ BUG_ON(q-type != C2_MQ_ADAPTER_TARGET);
+
+ while (wqe_count--) {
+ BUG_ON(c2_mq_empty(q));
+ *q-shared = cpu_to_be16((be16_to_cpu(*q-shared)+1) %
q-q_size);
+ }
+}
+
+
+u32 c2_mq_count(struct c2_mq *q)
+{
+ s32 count;
+
+ if (q-type == C2_MQ_HOST_TARGET) {
+ count = be16_to_cpu(*q-shared) - q-priv;
+ } else {
+ count = q-priv - be16_to_cpu(*q-shared);
+ }
+
+ if (count 0) {
+ count += q-q_size;
+ }
+
+ return (u32) count;
+}
+
+void c2_mq_req_init(struct c2_mq *q, u32 index, u32 q_size, u32 msg_size,
+ u8 __iomem *pool_start, u16 __iomem *peer, u32 type)
[PATCH v2 4/7] AMSO1100 Memory Management.
Review Changes:
- sizeof - sizeof()
---
drivers/infiniband/hw/amso1100/c2_alloc.c | 256
drivers/infiniband/hw/amso1100/c2_mm.c| 378 +
2 files changed, 634 insertions(+), 0 deletions(-)
diff --git a/drivers/infiniband/hw/amso1100/c2_alloc.c
b/drivers/infiniband/hw/amso1100/c2_alloc.c
new file mode 100644
index 000..e496eb7
--- /dev/null
+++ b/drivers/infiniband/hw/amso1100/c2_alloc.c
@@ -0,0 +1,256 @@
+/*
+ * Copyright (c) 2004 Topspin Communications. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer in the documentation and/or other materials
+ *provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include linux/errno.h
+#include linux/slab.h
+#include linux/bitmap.h
+
+#include c2.h
+
+/* Trivial bitmap-based allocator */
+u32 c2_alloc(struct c2_alloc *alloc)
+{
+ u32 obj;
+
+ spin_lock(alloc-lock);
+ obj = find_next_zero_bit(alloc-table, alloc-max, alloc-last);
+ if (obj = alloc-max)
+ obj = find_first_zero_bit(alloc-table, alloc-max);
+ if (obj = 0) {
+ alloc-last = obj+1;
+ if (alloc-last alloc-max)
+ alloc-last = 0;
+ }
+ spin_unlock(alloc-lock);
+
+ return obj;
+}
+
+void c2_free(struct c2_alloc *alloc, u32 obj)
+{
+ spin_lock(alloc-lock);
+ clear_bit(obj, alloc-table);
+ spin_unlock(alloc-lock);
+}
+
+int c2_alloc_init(struct c2_alloc *alloc, u32 num, u32 reserved)
+{
+ int i;
+
+ alloc-last = 0;
+ alloc-max = num;
+ spin_lock_init(alloc-lock);
+ alloc-table = kmalloc(BITS_TO_LONGS(num) * sizeof(long), GFP_KERNEL);
+ if (!alloc-table)
+ return -ENOMEM;
+
+ bitmap_zero(alloc-table, num);
+ for (i = 0; i reserved; ++i)
+ set_bit(i, alloc-table);
+
+ return 0;
+}
+
+void c2_alloc_cleanup(struct c2_alloc *alloc)
+{
+ kfree(alloc-table);
+}
+
+/*
+ * Array of pointers with lazy allocation of leaf pages. Callers of
+ * _get, _set and _clear methods must use a lock or otherwise
+ * serialize access to the array.
+ */
+
+void *c2_array_get(struct c2_array *array, int index)
+{
+ int p = (index * sizeof(void *)) PAGE_SHIFT;
+
+ if (array-page_list[p].page) {
+ int i = index (PAGE_SIZE / sizeof(void *) - 1);
+ return array-page_list[p].page[i];
+ } else
+ return NULL;
+}
+
+int c2_array_set(struct c2_array *array, int index, void *value)
+{
+ int p = (index * sizeof(void *)) PAGE_SHIFT;
+
+ /* Allocate with GFP_ATOMIC because we'll be called with locks held. */
+ if (!array-page_list[p].page)
+ array-page_list[p].page =
+ (void **) get_zeroed_page(GFP_ATOMIC);
+
+ if (!array-page_list[p].page)
+ return -ENOMEM;
+
+ array-page_list[p].page[index (PAGE_SIZE / sizeof(void *) - 1)] =
+ value;
+ ++array-page_list[p].used;
+
+ return 0;
+}
+
+void c2_array_clear(struct c2_array *array, int index)
+{
+ int p = (index * sizeof(void *)) PAGE_SHIFT;
+
+ if (--array-page_list[p].used == 0) {
+ free_page((unsigned long) array-page_list[p].page);
+ array-page_list[p].page = NULL;
+ }
+
+ if (array-page_list[p].used 0)
+ pr_debug(Array %p index %d page %d with ref count %d 0\n,
+array, index, p, array-page_list[p].used);
+}
+
+int c2_array_init(struct c2_array *array, int nent)
+{
+ int npage = (nent * sizeof(void *) + PAGE_SIZE - 1) / PAGE_SIZE;
+ int i;
+
+
[PATCH v2 6/7] AMSO1100: Privileged Verbs Queues.
Review Changes:
dprintk() - pr_debug()
---
drivers/infiniband/hw/amso1100/c2_vq.c | 260
drivers/infiniband/hw/amso1100/c2_vq.h | 63
2 files changed, 323 insertions(+), 0 deletions(-)
diff --git a/drivers/infiniband/hw/amso1100/c2_vq.c
b/drivers/infiniband/hw/amso1100/c2_vq.c
new file mode 100644
index 000..445b1ed
--- /dev/null
+++ b/drivers/infiniband/hw/amso1100/c2_vq.c
@@ -0,0 +1,260 @@
+/*
+ * Copyright (c) 2005 Ammasso, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer in the documentation and/or other materials
+ *provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include linux/slab.h
+#include linux/spinlock.h
+
+#include c2_vq.h
+#include c2_provider.h
+
+/*
+ * Verbs Request Objects:
+ *
+ * VQ Request Objects are allocated by the kernel verbs handlers.
+ * They contain a wait object, a refcnt, an atomic bool indicating that the
+ * adapter has replied, and a copy of the verb reply work request.
+ * A pointer to the VQ Request Object is passed down in the context
+ * field of the work request message, and reflected back by the adapter
+ * in the verbs reply message. The function handle_vq() in the interrupt
+ * path will use this pointer to:
+ * 1) append a copy of the verbs reply message
+ * 2) mark that the reply is ready
+ * 3) wake up the kernel verbs handler blocked awaiting the reply.
+ *
+ *
+ * The kernel verbs handlers do a get to put a 2nd reference on the
+ * VQ Request object. If the kernel verbs handler exits before the adapter
+ * can respond, this extra reference will keep the VQ Request object around
+ * until the adapter's reply can be processed. The reason we need this is
+ * because a pointer to this object is stuffed into the context field of
+ * the verbs work request message, and reflected back in the reply message.
+ * It is used in the interrupt handler (handle_vq()) to wake up the appropriate
+ * kernel verb handler that is blocked awaiting the verb reply.
+ * So handle_vq() will do a put on the object when it's done accessing it.
+ * NOTE: If we guarantee that the kernel verb handler will never bail before
+ *getting the reply, then we don't need these refcnts.
+ *
+ *
+ * VQ Request objects are freed by the kernel verbs handlers only
+ * after the verb has been processed, or when the adapter fails and
+ * does not reply.
+ *
+ *
+ * Verbs Reply Buffers:
+ *
+ * VQ Reply bufs are local host memory copies of a
+ * outstanding Verb Request reply
+ * message. The are always allocated by the kernel verbs handlers, and _may_
be
+ * freed by either the kernel verbs handler -or- the interrupt handler. The
+ * kernel verbs handler _must_ free the repbuf, then free the vq request object
+ * in that order.
+ */
+
+int vq_init(struct c2_dev *c2dev)
+{
+ sprintf(c2dev-vq_cache_name, c2-vq:dev%c,
+ (char) ('0' + c2dev-devnum));
+ c2dev-host_msg_cache =
+ kmem_cache_create(c2dev-vq_cache_name, c2dev-rep_vq.msg_size, 0,
+ SLAB_HWCACHE_ALIGN, NULL, NULL);
+ if (c2dev-host_msg_cache == NULL) {
+ return -ENOMEM;
+ }
+ return 0;
+}
+
+void vq_term(struct c2_dev *c2dev)
+{
+ kmem_cache_destroy(c2dev-host_msg_cache);
+}
+
+/* vq_req_alloc - allocate a VQ Request Object and initialize it.
+ * The refcnt is set to 1.
+ */
+struct c2_vq_req *vq_req_alloc(struct c2_dev *c2dev)
+{
+ struct c2_vq_req *r;
+
+ r = kmalloc(sizeof(struct c2_vq_req), GFP_KERNEL);
+ if (r) {
+ init_waitqueue_head(r-wait_object);
+ r-reply_msg = (u64) NULL;
+
[PATCH v2 1/2] iWARP Connection Manager.
This patch provides the new files implementing the iWARP Connection
Manager.
Review Changes:
- sizeof - sizeof()
- removed printks
- removed TT debug code
- cleaned up lock/unlock around switch statements.
- waitqueue - completion for destroy path.
---
drivers/infiniband/core/iwcm.c | 877
include/rdma/iw_cm.h | 254
include/rdma/iw_cm_private.h | 62 +++
3 files changed, 1193 insertions(+), 0 deletions(-)
diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c
new file mode 100644
index 000..994bc79
--- /dev/null
+++ b/drivers/infiniband/core/iwcm.c
@@ -0,0 +1,877 @@
+/*
+ * Copyright (c) 2004, 2005 Intel Corporation. All rights reserved.
+ * Copyright (c) 2004 Topspin Corporation. All rights reserved.
+ * Copyright (c) 2004, 2005 Voltaire Corporation. All rights reserved.
+ * Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ * Copyright (c) 2005 Network Appliance, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ *copyright notice, this list of conditions and the following
+ *disclaimer in the documentation and/or other materials
+ *provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED AS IS, WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+#include linux/dma-mapping.h
+#include linux/err.h
+#include linux/idr.h
+#include linux/interrupt.h
+#include linux/pci.h
+#include linux/rbtree.h
+#include linux/spinlock.h
+#include linux/workqueue.h
+#include linux/completion.h
+#include rdma/iw_cm.h
+#include rdma/iw_cm_private.h
+#include rdma/ib_addr.h
+
+MODULE_AUTHOR(Tom Tucker);
+MODULE_DESCRIPTION(iWARP CM);
+MODULE_LICENSE(Dual BSD/GPL);
+
+static struct workqueue_struct *iwcm_wq;
+struct iwcm_work {
+ struct work_struct work;
+ struct iwcm_id_private *cm_id;
+ struct list_head list;
+ struct iw_cm_event event;
+};
+
+/*
+ * Release a reference on cm_id. If the last reference is being removed
+ * and iw_destroy_cm_id is waiting, wake up the waiting thread.
+ */
+static int iwcm_deref_id(struct iwcm_id_private *cm_id_priv)
+{
+ int ret = 0;
+
+ BUG_ON(atomic_read(cm_id_priv-refcount)==0);
+ if (atomic_dec_and_test(cm_id_priv-refcount)) {
+ BUG_ON(!list_empty(cm_id_priv-work_list));
+ if (waitqueue_active(cm_id_priv-destroy_comp.wait)) {
+ BUG_ON(cm_id_priv-state != IW_CM_STATE_DESTROYING);
+ BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY,
+ cm_id_priv-flags));
+ ret = 1;
+ }
+ complete(cm_id_priv-destroy_comp);
+ }
+
+ return ret;
+}
+
+static void add_ref(struct iw_cm_id *cm_id)
+{
+ struct iwcm_id_private *cm_id_priv;
+ cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+ atomic_inc(cm_id_priv-refcount);
+}
+
+static void rem_ref(struct iw_cm_id *cm_id)
+{
+ struct iwcm_id_private *cm_id_priv;
+ cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+ iwcm_deref_id(cm_id_priv);
+}
+
+static void cm_event_handler(struct iw_cm_id *cm_id, struct iw_cm_event
*event);
+
+struct iw_cm_id *iw_create_cm_id(struct ib_device *device,
+iw_cm_handler cm_handler,
+void *context)
+{
+ struct iwcm_id_private *cm_id_priv;
+
+ cm_id_priv = kzalloc(sizeof(*cm_id_priv), GFP_KERNEL);
+ if (!cm_id_priv)
+ return ERR_PTR(-ENOMEM);
+
+ cm_id_priv-state = IW_CM_STATE_IDLE;
+ cm_id_priv-id.device = device;
+ cm_id_priv-id.cm_handler = cm_handler;
+ cm_id_priv-id.context = context;
+ cm_id_priv-id.event_handler = cm_event_handler;
+
[PATCH v2 2/7] AMSO1100 WR / Event Definitions.
Resending 2/7 gziped. linux-kernel and netdev mailing lists didn't formward the plain text patch... If anywone knows how to address this issue, please email me directly cuz I don't know why 2/7 didn't get forwarded. Sorry. Steve. amso1100_wr.gz Description: GNU Zip compressed data
Re: [PATCH v2 2/7] AMSO1100 WR / Event Definitions.
I just realized it could be the spam filters. You have some comments with three 'X's in a row which might be getting it blocked. Is that possible? - R. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Updated sysctl documentation take #2
El Wed, 7 Jun 2006 12:46:41 -0700, Stephen Hemminger [EMAIL PROTECTED] escribió: The network stuff was all in Documentation/networking/ip-sysctl.txt. Someone else has already started fixing it. Darn, didn't know that. Anyway, I merged that file into Documentation/sysctl/ because IMO all sysctls must be in the same place. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH v2 2/7] AMSO1100 WR / Event Definitions.
On Wed, 2006-06-07 at 13:43 -0700, Roland Dreier wrote: I just realized it could be the spam filters. You have some comments with three 'X's in a row which might be getting it blocked. Is that possible? There are other files that have comments with 'XXX' like c2_provider.c and c2_qp.c which is in patch 3/7 and it made it though. These 'XXX' comments need to be cleaned up anyway, so I'll remove them (or address the issue if there is one) and we'll see next time I post a new version. Steve. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] pcnet32 driver NAPI support
On Wed, Jun 07, 2006 at 03:34:56PM -0400, Lennart Sorensen wrote: On Wed, Jun 07, 2006 at 11:20:40AM -0700, Don Fry wrote: Some areas of concern that you may have addressed already, I have not scanned your changes yet, are what happens if the ring size is changed without bringing down the interface (via ethtool), or if the loopback test is run in a similar fashion, or a tx timeout occurs. The same thing as if it was done before enabling napi. From a few messages I have seen, it doesn't work right now, and it won't work any better with my changes. I have never tried changing the ring size on the fly, so I don't know. Today without your NAPI changes the device can be up and operational, and change the ring size without hanging or causing a panic or causing problems with the driver. The same can be said for running the loopback test, or when a tx timeout occurs. It does not look like the same can be said for your NAPI changes, yet. Try changing the ring size a dozen times during a receive storm and see what happens. It appears that the port is stopped before the ring size change is done, although I can't really tell how it handles things if the queue is not empty when it stops the port. Does it try to handle anything left in the ring first or does it just toss those packets? (That I would consider wrong). The lp-lock MUST be held whenever accessing the csr or bcr registers as this is a multi-step process, and has been the source of problems in the past. Even on UP systems. Hmm, I just followed what appeared to be in pcnet32_rx and how tulip and a few other drivers had done their napi conversions. It certainly works for me the way I did it. Haven't seen any lockups yet. I do see that I am not holding the lock when I acknowledge IRQs in pcnet32_poll, which pcnet32_rx doesn't need to worry about since it is called from the interrupt handler which already holds the lock. That should be fixed then. Yes, that is the minimum that needs to be changed. There have been a lot of changes made to the driver, by myself included, that seem to work just fine, but later the timing changes and you lose the race, resulting in problems. Changing one part of the driver without understanding the whole thing is very risky. So I can do: // Clear RX interrupts spin_lock(lp-lock); lp-a.write_csr (ioaddr, 0, 0x1400); spin_unlock(lp-lock); That part seems simple enough to protect. Is this safe without holding the lock? } while(lp-a.read_csr (ioaddr, 0) 0x1400); No, see below. Not sure how to wrap a lock around that one without holding the lock for way too long. perhaps: spin_lock(lp-lock); state=lp-a.read_csr (ioaddr, 0) 0x1400; spin_unlock(lp-lock); } while(state); Does that seem more reasonable? The lock must be held during ANY read or write to a csr or bcr. The problem is that accessing a csr/bcr takes two steps. One is to write the address register indicating which register to read or write, and second, read or write the register. The problem is that without the lock, entity A writes to the rap to access register X, then before it can access that register entity B writes the rap to access register Y, At that point entity A will read, or worse write, the contents of the wrong register. This problem is even worse when reading/writing a PHY register because it entails writing bcr 33 and reading/writing bcr 34, without being interrupted by something that might change bcr 33. Not likely, but not something that can be ignored. It may not happen often, but it will happen eventually. Len Sorensen -- Don Fry [EMAIL PROTECTED] - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] pcnet32 driver NAPI support
One other problem I ran into. I applied the patch but it will not compile because rl_active is never defined. I have worked around it but ... -- Don Fry [EMAIL PROTECTED] - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Updated sysctl documentation take #2
El Wed, 7 Jun 2006 13:06:53 -0700, Randy.Dunlap [EMAIL PROTECTED] escribió: I don't know how long it takes to review such a large patch, but I'll continue to do so. For now: Yeah, my english is poor ;) Most of the sysctl documentation cames from other files anyway, README was (re)written by me, so the rest is not so bad... OK, that's all for the README file. I'll look at the rest of it sometime this week. I don't think that it's quite ready to be merged. Thank's for your review, altought I didn't though someone was to review so deeply a documentation patch ;) I've gone through all the files and fixed the 72-col limit and everything I could. I've updated the patch http://terra.es/personal/diegocg/sysctl-docs - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 9/10] Add Vitesse 8244 PHY for MPC8641 HPCN platform.
Signed-off-by: Kriston Carson [EMAIL PROTECTED]
Signed-off-by: Xianghua Xiao [EMAIL PROTECTED]
Signed-off-by: Jon Loeliger [EMAIL PROTECTED]
---
drivers/net/Kconfig |6 +-
drivers/net/phy/Kconfig |5 ++
drivers/net/phy/Makefile |1
drivers/net/phy/vitesse.c | 136 +
4 files changed, 145 insertions(+), 3 deletions(-)
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index bdaaad8..c1c2758 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2179,11 +2179,11 @@ config SPIDER_NET
config GIANFAR
tristate Gianfar Ethernet
- depends on 85xx || 83xx
+ depends on 85xx || 83xx || PPC_86xx
select PHYLIB
help
- This driver supports the Gigabit TSEC on the MPC85xx
- family of chips, and the FEC on the 8540
+ This driver supports the Gigabit TSEC on the MPC83xx, MPC85xx,
+ and MPC86xx family of chips, and the FEC on the 8540.
config GFAR_NAPI
bool NAPI Support
diff --git a/drivers/net/phy/Kconfig b/drivers/net/phy/Kconfig
index fa39b94..76e51b1 100644
--- a/drivers/net/phy/Kconfig
+++ b/drivers/net/phy/Kconfig
@@ -44,6 +44,11 @@ config CICADA_PHY
depends on PHYLIB
---help---
Currently supports the cis8204
+config VITESSE_PHY
+tristate Drivers for the Vitesse PHYs
+depends on PHYLIB
+---help---
+ Currently supports the vsc8244
endmenu
diff --git a/drivers/net/phy/Makefile b/drivers/net/phy/Makefile
index e4116a5..a8d066e 100644
--- a/drivers/net/phy/Makefile
+++ b/drivers/net/phy/Makefile
@@ -8,3 +8,4 @@ obj-$(CONFIG_DAVICOM_PHY) += davicom.o
obj-$(CONFIG_CICADA_PHY) += cicada.o
obj-$(CONFIG_LXT_PHY) += lxt.o
obj-$(CONFIG_QSEMI_PHY)+= qsemi.o
+obj-$(CONFIG_VITESSE_PHY) += vitesse.o
diff --git a/drivers/net/phy/vitesse.c b/drivers/net/phy/vitesse.c
new file mode 100644
index 000..45a605f
--- /dev/null
+++ b/drivers/net/phy/vitesse.c
@@ -0,0 +1,136 @@
+/*
+ * Driver for Vitesse PHYs
+ *
+ * Author: Kriston Carson
+ *
+ * Copyright (c) 2005 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ */
+
+#include linux/config.h
+#include linux/kernel.h
+#include linux/sched.h
+#include linux/string.h
+#include linux/errno.h
+#include linux/unistd.h
+#include linux/slab.h
+#include linux/interrupt.h
+#include linux/init.h
+#include linux/delay.h
+#include linux/netdevice.h
+#include linux/etherdevice.h
+#include linux/skbuff.h
+#include linux/spinlock.h
+#include linux/mm.h
+#include linux/module.h
+#include linux/version.h
+#include linux/mii.h
+#include linux/ethtool.h
+#include linux/phy.h
+
+#include asm/io.h
+#include asm/irq.h
+#include asm/uaccess.h
+
+/* Vitesse Extended Control Register 1 */
+#define MII_VSC8244_EXT_CON1 0x17
+#define MII_VSC8244_EXTCON1_INIT 0x
+
+/* Vitesse Interrupt Mask Register */
+#define MII_VSC8244_IMASK 0x19
+#define MII_VSC8244_IMASK_IEN 0x8000
+#define MII_VSC8244_IMASK_SPEED0x4000
+#define MII_VSC8244_IMASK_LINK 0x2000
+#define MII_VSC8244_IMASK_DUPLEX 0x1000
+#define MII_VSC8244_IMASK_MASK 0xf000
+
+/* Vitesse Interrupt Status Register */
+#define MII_VSC8244_ISTAT 0x1a
+#define MII_VSC8244_ISTAT_STATUS 0x8000
+#define MII_VSC8244_ISTAT_SPEED0x4000
+#define MII_VSC8244_ISTAT_LINK 0x2000
+#define MII_VSC8244_ISTAT_DUPLEX 0x1000
+
+/* Vitesse Auxiliary Control/Status Register */
+#define MII_VSC8244_AUX_CONSTAT0x1c
+#define MII_VSC8244_AUXCONSTAT_INIT0x0004
+#define MII_VSC8244_AUXCONSTAT_DUPLEX 0x0020
+#define MII_VSC8244_AUXCONSTAT_SPEED 0x0018
+#define MII_VSC8244_AUXCONSTAT_GBIT0x0010
+#define MII_VSC8244_AUXCONSTAT_100 0x0008
+
+MODULE_DESCRIPTION(Vitesse PHY driver);
+MODULE_AUTHOR(Kriston Carson);
+MODULE_LICENSE(GPL);
+
+static int vsc824x_config_init(struct phy_device *phydev)
+{
+ int err;
+
+ err = phy_write(phydev, MII_VSC8244_AUX_CONSTAT,
+ MII_VSC8244_AUXCONSTAT_INIT);
+
+ if (err 0)
+ return err;
+
+ err = phy_write(phydev, MII_VSC8244_EXT_CON1,
+ MII_VSC8244_EXTCON1_INIT);
+
+ return err;
+
+}
+
+static int vsc824x_ack_interrupt(struct phy_device *phydev)
+{
+ int err = phy_read(phydev, MII_VSC8244_ISTAT);
+
+ return (err 0) ? err : 0;
+
+}
+
+static int vsc824x_config_intr(struct phy_device *phydev)
+{
+ int err;
+
+ if (phydev-interrupts == PHY_INTERRUPT_ENABLED)
+ err = phy_write(phydev, MII_VSC8244_IMASK,
+
Re: [PATCH v2 2/2] iWARP Core Changes.
A reference is being taken on an iWARP device that is never getting
released. This prevents a participating iWARP netdev device from being
unloaded after a connection has been established on the passive side.
Search for ip_dev_find below...
On Wed, 2006-06-07 at 15:06 -0500, Steve Wise wrote:
This patch contains modifications to the existing rdma header files,
core files, drivers, and ulp files to support iWARP.
Review updates:
- copy_addr() - rdma_copy_addr()
- dst_dev_addr param in rdma_copy_addr to const.
- various spacing nits with recasting
- include linux/inetdevice.h to get ip_dev_find() prototype.
---
drivers/infiniband/core/Makefile |4
drivers/infiniband/core/addr.c | 19 +
drivers/infiniband/core/cache.c |8 -
drivers/infiniband/core/cm.c |3
drivers/infiniband/core/cma.c| 353
+++---
drivers/infiniband/core/device.c |6
drivers/infiniband/core/mad.c| 11 +
drivers/infiniband/core/sa_query.c |5
drivers/infiniband/core/smi.c| 18 +
drivers/infiniband/core/sysfs.c | 18 +
drivers/infiniband/core/ucm.c|5
drivers/infiniband/core/user_mad.c |9 -
drivers/infiniband/hw/ipath/ipath_verbs.c|2
drivers/infiniband/hw/mthca/mthca_provider.c |2
drivers/infiniband/ulp/ipoib/ipoib_main.c|8 +
drivers/infiniband/ulp/srp/ib_srp.c |2
include/rdma/ib_addr.h | 15 +
include/rdma/ib_verbs.h | 39 +++
18 files changed, 435 insertions(+), 92 deletions(-)
diff --git a/drivers/infiniband/core/Makefile
b/drivers/infiniband/core/Makefile
index 68e73ec..163d991 100644
--- a/drivers/infiniband/core/Makefile
+++ b/drivers/infiniband/core/Makefile
@@ -1,7 +1,7 @@
infiniband-$(CONFIG_INFINIBAND_ADDR_TRANS) := ib_addr.o rdma_cm.o
obj-$(CONFIG_INFINIBAND) += ib_core.o ib_mad.o ib_sa.o \
- ib_cm.o $(infiniband-y)
+ ib_cm.o iw_cm.o $(infiniband-y)
obj-$(CONFIG_INFINIBAND_USER_MAD) += ib_umad.o
obj-$(CONFIG_INFINIBAND_USER_ACCESS) += ib_uverbs.o ib_ucm.o
@@ -14,6 +14,8 @@ ib_sa-y := sa_query.o
ib_cm-y := cm.o
+iw_cm-y := iwcm.o
+
rdma_cm-y := cma.o
ib_addr-y := addr.o
diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c
index d294bbc..83f84ef 100644
--- a/drivers/infiniband/core/addr.c
+++ b/drivers/infiniband/core/addr.c
@@ -32,6 +32,7 @@ #include linux/mutex.h
#include linux/inetdevice.h
#include linux/workqueue.h
#include linux/if_arp.h
+#include linux/inetdevice.h
#include net/arp.h
#include net/neighbour.h
#include net/route.h
@@ -60,12 +61,15 @@ static LIST_HEAD(req_list);
static DECLARE_WORK(work, process_req, NULL);
static struct workqueue_struct *addr_wq;
-static int copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
- unsigned char *dst_dev_addr)
+int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
+ const unsigned char *dst_dev_addr)
{
switch (dev-type) {
case ARPHRD_INFINIBAND:
- dev_addr-dev_type = IB_NODE_CA;
+ dev_addr-dev_type = RDMA_NODE_IB_CA;
+ break;
+ case ARPHRD_ETHER:
+ dev_addr-dev_type = RDMA_NODE_RNIC;
break;
default:
return -EADDRNOTAVAIL;
@@ -77,6 +81,7 @@ static int copy_addr(struct rdma_dev_add
memcpy(dev_addr-dst_dev_addr, dst_dev_addr, MAX_ADDR_LEN);
return 0;
}
+EXPORT_SYMBOL(rdma_copy_addr);
int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr)
{
@@ -88,7 +93,7 @@ int rdma_translate_ip(struct sockaddr *a
if (!dev)
return -EADDRNOTAVAIL;
- ret = copy_addr(dev_addr, dev, NULL);
+ ret = rdma_copy_addr(dev_addr, dev, NULL);
dev_put(dev);
return ret;
}
@@ -160,7 +165,7 @@ static int addr_resolve_remote(struct so
/* If the device does ARP internally, return 'done' */
if (rt-idev-dev-flags IFF_NOARP) {
- copy_addr(addr, rt-idev-dev, NULL);
+ rdma_copy_addr(addr, rt-idev-dev, NULL);
goto put;
}
@@ -180,7 +185,7 @@ static int addr_resolve_remote(struct so
src_in-sin_addr.s_addr = rt-rt_src;
}
- ret = copy_addr(addr, neigh-dev, neigh-ha);
+ ret = rdma_copy_addr(addr, neigh-dev, neigh-ha);
release:
neigh_release(neigh);
put:
@@ -244,7 +249,7 @@ static int addr_resolve_local(struct soc
if (ZERONET(src_ip)) {
src_in-sin_family = dst_in-sin_family;
Re: Updated sysctl documentation take #2
On Wed, Jun 07, 2006, Diego Calleja wrote: Since people didn't like the many small files approach, I've moved it to directories containing index.txt files: Documentation/sysctl/vm/index.txt Documentation/sysctl/net/core/index.txt Why not just Documentation/sysctl/vm.txt Documentation/sysctl/net/core.txt etc.? Johannes - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Updated sysctl documentation take #2
On Thu, 8 Jun 2006, Johannes Stezenbach wrote: On Wed, Jun 07, 2006, Diego Calleja wrote: Since people didn't like the many small files approach, I've moved it to directories containing index.txt files: Documentation/sysctl/vm/index.txt Documentation/sysctl/net/core/index.txt Why not just Documentation/sysctl/vm.txt Documentation/sysctl/net/core.txt etc.? better matching of the proc layout would be my guess. - To unsubscribe from this list: send the line unsubscribe netdev in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
