linux_dsm_epyc7002/drivers/net/chelsio/pm3393.c

797 lines
30 KiB
C
Raw Normal View History

[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
/*****************************************************************************
* *
* File: pm3393.c *
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
* $Revision: 1.16 $ *
* $Date: 2005/05/14 00:59:32 $ *
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
* Description: *
* PMC/SIERRA (pm3393) MAC-PHY functionality. *
* part of the Chelsio 10Gb Ethernet Driver. *
* *
* This program is free software; you can redistribute it and/or modify *
* it under the terms of the GNU General Public License, version 2, as *
* published by the Free Software Foundation. *
* *
* 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. *
* *
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED *
* WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF *
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. *
* *
* http://www.chelsio.com *
* *
* Copyright (c) 2003 - 2005 Chelsio Communications, Inc. *
* All rights reserved. *
* *
* Maintainers: maintainers@chelsio.com *
* *
* Authors: Dimitrios Michailidis <dm@chelsio.com> *
* Tina Yang <tainay@chelsio.com> *
* Felix Marti <felix@chelsio.com> *
* Scott Bardone <sbardone@chelsio.com> *
* Kurt Ottaway <kottaway@chelsio.com> *
* Frank DiMambro <frank@chelsio.com> *
* *
* History: *
* *
****************************************************************************/
#include "common.h"
#include "regs.h"
#include "gmac.h"
#include "elmer0.h"
#include "suni1x10gexp_regs.h"
#include <linux/crc32.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 15:04:11 +07:00
#include <linux/slab.h>
#define OFFSET(REG_ADDR) ((REG_ADDR) << 2)
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
/* Max frame size PM3393 can handle. Includes Ethernet header and CRC. */
#define MAX_FRAME_SIZE 9600
#define IPG 12
#define TXXG_CONF1_VAL ((IPG << SUNI1x10GEXP_BITOFF_TXXG_IPGT) | \
SUNI1x10GEXP_BITMSK_TXXG_32BIT_ALIGN | SUNI1x10GEXP_BITMSK_TXXG_CRCEN | \
SUNI1x10GEXP_BITMSK_TXXG_PADEN)
#define RXXG_CONF1_VAL (SUNI1x10GEXP_BITMSK_RXXG_PUREP | 0x14 | \
SUNI1x10GEXP_BITMSK_RXXG_FLCHK | SUNI1x10GEXP_BITMSK_RXXG_CRC_STRIP)
/* Update statistics every 15 minutes */
#define STATS_TICK_SECS (15 * 60)
enum { /* RMON registers */
RxOctetsReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_1_LOW,
RxUnicastFramesReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_4_LOW,
RxMulticastFramesReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_5_LOW,
RxBroadcastFramesReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_6_LOW,
RxPAUSEMACCtrlFramesReceived = SUNI1x10GEXP_REG_MSTAT_COUNTER_8_LOW,
RxFrameCheckSequenceErrors = SUNI1x10GEXP_REG_MSTAT_COUNTER_10_LOW,
RxFramesLostDueToInternalMACErrors = SUNI1x10GEXP_REG_MSTAT_COUNTER_11_LOW,
RxSymbolErrors = SUNI1x10GEXP_REG_MSTAT_COUNTER_12_LOW,
RxInRangeLengthErrors = SUNI1x10GEXP_REG_MSTAT_COUNTER_13_LOW,
RxFramesTooLongErrors = SUNI1x10GEXP_REG_MSTAT_COUNTER_15_LOW,
RxJabbers = SUNI1x10GEXP_REG_MSTAT_COUNTER_16_LOW,
RxFragments = SUNI1x10GEXP_REG_MSTAT_COUNTER_17_LOW,
RxUndersizedFrames = SUNI1x10GEXP_REG_MSTAT_COUNTER_18_LOW,
RxJumboFramesReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_25_LOW,
RxJumboOctetsReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_26_LOW,
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
TxOctetsTransmittedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_33_LOW,
TxFramesLostDueToInternalMACTransmissionError = SUNI1x10GEXP_REG_MSTAT_COUNTER_35_LOW,
TxTransmitSystemError = SUNI1x10GEXP_REG_MSTAT_COUNTER_36_LOW,
TxUnicastFramesTransmittedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_38_LOW,
TxMulticastFramesTransmittedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_40_LOW,
TxBroadcastFramesTransmittedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_42_LOW,
TxPAUSEMACCtrlFramesTransmitted = SUNI1x10GEXP_REG_MSTAT_COUNTER_43_LOW,
TxJumboFramesReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_51_LOW,
TxJumboOctetsReceivedOK = SUNI1x10GEXP_REG_MSTAT_COUNTER_52_LOW
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
};
struct _cmac_instance {
u8 enabled;
u8 fc;
u8 mac_addr[6];
};
static int pmread(struct cmac *cmac, u32 reg, u32 * data32)
{
t1_tpi_read(cmac->adapter, OFFSET(reg), data32);
return 0;
}
static int pmwrite(struct cmac *cmac, u32 reg, u32 data32)
{
t1_tpi_write(cmac->adapter, OFFSET(reg), data32);
return 0;
}
/* Port reset. */
static int pm3393_reset(struct cmac *cmac)
{
return 0;
}
/*
* Enable interrupts for the PM3393
*
* 1. Enable PM3393 BLOCK interrupts.
* 2. Enable PM3393 Master Interrupt bit(INTE)
* 3. Enable ELMER's PM3393 bit.
* 4. Enable Terminator external interrupt.
*/
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
static int pm3393_interrupt_enable(struct cmac *cmac)
{
u32 pl_intr;
/* PM3393 - Enabling all hardware block interrupts.
*/
pmwrite(cmac, SUNI1x10GEXP_REG_SERDES_3125_INTERRUPT_ENABLE, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_XRF_INTERRUPT_ENABLE, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_XRF_DIAG_INTERRUPT_ENABLE, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_RXOAM_INTERRUPT_ENABLE, 0xffff);
/* Don't interrupt on statistics overflow, we are polling */
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_0, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_1, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_2, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_3, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_IFLX_FIFO_OVERFLOW_ENABLE, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_PL4ODP_INTERRUPT_MASK, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_XTEF_INTERRUPT_ENABLE, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_TXOAM_INTERRUPT_ENABLE, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_CONFIG_3, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_PL4IO_LOCK_DETECT_MASK, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_CONFIG_3, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_PL4IDU_INTERRUPT_MASK, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_EFLX_FIFO_OVERFLOW_ERROR_ENABLE, 0xffff);
/* PM3393 - Global interrupt enable
*/
/* TBD XXX Disable for now until we figure out why error interrupts keep asserting. */
pmwrite(cmac, SUNI1x10GEXP_REG_GLOBAL_INTERRUPT_ENABLE,
0 /*SUNI1x10GEXP_BITMSK_TOP_INTE */ );
/* TERMINATOR - PL_INTERUPTS_EXT */
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
pl_intr = readl(cmac->adapter->regs + A_PL_ENABLE);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
pl_intr |= F_PL_INTR_EXT;
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
writel(pl_intr, cmac->adapter->regs + A_PL_ENABLE);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
return 0;
}
static int pm3393_interrupt_disable(struct cmac *cmac)
{
u32 elmer;
/* PM3393 - Enabling HW interrupt blocks. */
pmwrite(cmac, SUNI1x10GEXP_REG_SERDES_3125_INTERRUPT_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_XRF_INTERRUPT_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_XRF_DIAG_INTERRUPT_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_RXOAM_INTERRUPT_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_0, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_1, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_2, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_INTERRUPT_MASK_3, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_IFLX_FIFO_OVERFLOW_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_PL4ODP_INTERRUPT_MASK, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_XTEF_INTERRUPT_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_TXOAM_INTERRUPT_ENABLE, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_CONFIG_3, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_PL4IO_LOCK_DETECT_MASK, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_CONFIG_3, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_PL4IDU_INTERRUPT_MASK, 0);
pmwrite(cmac, SUNI1x10GEXP_REG_EFLX_FIFO_OVERFLOW_ERROR_ENABLE, 0);
/* PM3393 - Global interrupt enable */
pmwrite(cmac, SUNI1x10GEXP_REG_GLOBAL_INTERRUPT_ENABLE, 0);
/* ELMER - External chip interrupts. */
t1_tpi_read(cmac->adapter, A_ELMER0_INT_ENABLE, &elmer);
elmer &= ~ELMER0_GP_BIT1;
t1_tpi_write(cmac->adapter, A_ELMER0_INT_ENABLE, elmer);
/* TERMINATOR - PL_INTERUPTS_EXT */
/* DO NOT DISABLE TERMINATOR's EXTERNAL INTERRUPTS. ANOTHER CHIP
* COULD WANT THEM ENABLED. We disable PM3393 at the ELMER level.
*/
return 0;
}
static int pm3393_interrupt_clear(struct cmac *cmac)
{
u32 elmer;
u32 pl_intr;
u32 val32;
/* PM3393 - Clearing HW interrupt blocks. Note, this assumes
* bit WCIMODE=0 for a clear-on-read.
*/
pmread(cmac, SUNI1x10GEXP_REG_SERDES_3125_INTERRUPT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_XRF_INTERRUPT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_XRF_DIAG_INTERRUPT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_RXOAM_INTERRUPT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_PL4ODP_INTERRUPT, &val32);
pmread(cmac, SUNI1x10GEXP_REG_XTEF_INTERRUPT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_IFLX_FIFO_OVERFLOW_INTERRUPT, &val32);
pmread(cmac, SUNI1x10GEXP_REG_TXOAM_INTERRUPT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_RXXG_INTERRUPT, &val32);
pmread(cmac, SUNI1x10GEXP_REG_TXXG_INTERRUPT, &val32);
pmread(cmac, SUNI1x10GEXP_REG_PL4IDU_INTERRUPT, &val32);
pmread(cmac, SUNI1x10GEXP_REG_EFLX_FIFO_OVERFLOW_ERROR_INDICATION,
&val32);
pmread(cmac, SUNI1x10GEXP_REG_PL4IO_LOCK_DETECT_STATUS, &val32);
pmread(cmac, SUNI1x10GEXP_REG_PL4IO_LOCK_DETECT_CHANGE, &val32);
/* PM3393 - Global interrupt status
*/
pmread(cmac, SUNI1x10GEXP_REG_MASTER_INTERRUPT_STATUS, &val32);
/* ELMER - External chip interrupts.
*/
t1_tpi_read(cmac->adapter, A_ELMER0_INT_CAUSE, &elmer);
elmer |= ELMER0_GP_BIT1;
t1_tpi_write(cmac->adapter, A_ELMER0_INT_CAUSE, elmer);
/* TERMINATOR - PL_INTERUPTS_EXT
*/
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
pl_intr = readl(cmac->adapter->regs + A_PL_CAUSE);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
pl_intr |= F_PL_INTR_EXT;
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
writel(pl_intr, cmac->adapter->regs + A_PL_CAUSE);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
return 0;
}
/* Interrupt handler */
static int pm3393_interrupt_handler(struct cmac *cmac)
{
u32 master_intr_status;
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
/* Read the master interrupt status register. */
pmread(cmac, SUNI1x10GEXP_REG_MASTER_INTERRUPT_STATUS,
&master_intr_status);
if (netif_msg_intr(cmac->adapter))
dev_dbg(&cmac->adapter->pdev->dev, "PM3393 intr cause 0x%x\n",
master_intr_status);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
/* TBD XXX Lets just clear everything for now */
pm3393_interrupt_clear(cmac);
return 0;
}
static int pm3393_enable(struct cmac *cmac, int which)
{
if (which & MAC_DIRECTION_RX)
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_CONFIG_1,
(RXXG_CONF1_VAL | SUNI1x10GEXP_BITMSK_RXXG_RXEN));
if (which & MAC_DIRECTION_TX) {
u32 val = TXXG_CONF1_VAL | SUNI1x10GEXP_BITMSK_TXXG_TXEN0;
if (cmac->instance->fc & PAUSE_RX)
val |= SUNI1x10GEXP_BITMSK_TXXG_FCRX;
if (cmac->instance->fc & PAUSE_TX)
val |= SUNI1x10GEXP_BITMSK_TXXG_FCTX;
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_CONFIG_1, val);
}
cmac->instance->enabled |= which;
return 0;
}
static int pm3393_enable_port(struct cmac *cmac, int which)
{
/* Clear port statistics */
pmwrite(cmac, SUNI1x10GEXP_REG_MSTAT_CONTROL,
SUNI1x10GEXP_BITMSK_MSTAT_CLEAR);
udelay(2);
memset(&cmac->stats, 0, sizeof(struct cmac_statistics));
pm3393_enable(cmac, which);
/*
* XXX This should be done by the PHY and preferably not at all.
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
* The PHY doesn't give us link status indication on its own so have
* the link management code query it instead.
*/
t1_link_changed(cmac->adapter, 0);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
return 0;
}
static int pm3393_disable(struct cmac *cmac, int which)
{
if (which & MAC_DIRECTION_RX)
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_CONFIG_1, RXXG_CONF1_VAL);
if (which & MAC_DIRECTION_TX)
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_CONFIG_1, TXXG_CONF1_VAL);
/*
* The disable is graceful. Give the PM3393 time. Can't wait very
* long here, we may be holding locks.
*/
udelay(20);
cmac->instance->enabled &= ~which;
return 0;
}
static int pm3393_loopback_enable(struct cmac *cmac)
{
return 0;
}
static int pm3393_loopback_disable(struct cmac *cmac)
{
return 0;
}
static int pm3393_set_mtu(struct cmac *cmac, int mtu)
{
int enabled = cmac->instance->enabled;
/* MAX_FRAME_SIZE includes header + FCS, mtu doesn't */
mtu += 14 + 4;
if (mtu > MAX_FRAME_SIZE)
return -EINVAL;
/* Disable Rx/Tx MAC before configuring it. */
if (enabled)
pm3393_disable(cmac, MAC_DIRECTION_RX | MAC_DIRECTION_TX);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MAX_FRAME_LENGTH, mtu);
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_MAX_FRAME_SIZE, mtu);
if (enabled)
pm3393_enable(cmac, enabled);
return 0;
}
static int pm3393_set_rx_mode(struct cmac *cmac, struct t1_rx_mode *rm)
{
int enabled = cmac->instance->enabled & MAC_DIRECTION_RX;
u32 rx_mode;
/* Disable MAC RX before reconfiguring it */
if (enabled)
pm3393_disable(cmac, MAC_DIRECTION_RX);
pmread(cmac, SUNI1x10GEXP_REG_RXXG_ADDRESS_FILTER_CONTROL_2, &rx_mode);
rx_mode &= ~(SUNI1x10GEXP_BITMSK_RXXG_PMODE |
SUNI1x10GEXP_BITMSK_RXXG_MHASH_EN);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_ADDRESS_FILTER_CONTROL_2,
(u16)rx_mode);
if (t1_rx_mode_promisc(rm)) {
/* Promiscuous mode. */
rx_mode |= SUNI1x10GEXP_BITMSK_RXXG_PMODE;
}
if (t1_rx_mode_allmulti(rm)) {
/* Accept all multicast. */
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_LOW, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_MIDLOW, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_MIDHIGH, 0xffff);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_HIGH, 0xffff);
rx_mode |= SUNI1x10GEXP_BITMSK_RXXG_MHASH_EN;
} else if (t1_rx_mode_mc_cnt(rm)) {
/* Accept one or more multicast(s). */
struct netdev_hw_addr *ha;
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
int bit;
u16 mc_filter[4] = { 0, };
netdev_for_each_mc_addr(ha, t1_get_netdev(rm)) {
/* bit[23:28] */
bit = (ether_crc(ETH_ALEN, ha->addr) >> 23) & 0x3f;
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
mc_filter[bit >> 4] |= 1 << (bit & 0xf);
}
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_LOW, mc_filter[0]);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_MIDLOW, mc_filter[1]);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_MIDHIGH, mc_filter[2]);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_MULTICAST_HASH_HIGH, mc_filter[3]);
rx_mode |= SUNI1x10GEXP_BITMSK_RXXG_MHASH_EN;
}
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_ADDRESS_FILTER_CONTROL_2, (u16)rx_mode);
if (enabled)
pm3393_enable(cmac, MAC_DIRECTION_RX);
return 0;
}
static int pm3393_get_speed_duplex_fc(struct cmac *cmac, int *speed,
int *duplex, int *fc)
{
if (speed)
*speed = SPEED_10000;
if (duplex)
*duplex = DUPLEX_FULL;
if (fc)
*fc = cmac->instance->fc;
return 0;
}
static int pm3393_set_speed_duplex_fc(struct cmac *cmac, int speed, int duplex,
int fc)
{
if (speed >= 0 && speed != SPEED_10000)
return -1;
if (duplex >= 0 && duplex != DUPLEX_FULL)
return -1;
if (fc & ~(PAUSE_TX | PAUSE_RX))
return -1;
if (fc != cmac->instance->fc) {
cmac->instance->fc = (u8) fc;
if (cmac->instance->enabled & MAC_DIRECTION_TX)
pm3393_enable(cmac, MAC_DIRECTION_TX);
}
return 0;
}
#define RMON_UPDATE(mac, name, stat_name) \
{ \
t1_tpi_read((mac)->adapter, OFFSET(name), &val0); \
t1_tpi_read((mac)->adapter, OFFSET((name)+1), &val1); \
t1_tpi_read((mac)->adapter, OFFSET((name)+2), &val2); \
(mac)->stats.stat_name = (u64)(val0 & 0xffff) | \
((u64)(val1 & 0xffff) << 16) | \
((u64)(val2 & 0xff) << 32) | \
((mac)->stats.stat_name & \
0xffffff0000000000ULL); \
if (ro & \
(1ULL << ((name - SUNI1x10GEXP_REG_MSTAT_COUNTER_0_LOW) >> 2))) \
(mac)->stats.stat_name += 1ULL << 40; \
}
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
static const struct cmac_statistics *pm3393_update_statistics(struct cmac *mac,
int flag)
{
u64 ro;
u32 val0, val1, val2, val3;
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
/* Snap the counters */
pmwrite(mac, SUNI1x10GEXP_REG_MSTAT_CONTROL,
SUNI1x10GEXP_BITMSK_MSTAT_SNAP);
/* Counter rollover, clear on read */
pmread(mac, SUNI1x10GEXP_REG_MSTAT_COUNTER_ROLLOVER_0, &val0);
pmread(mac, SUNI1x10GEXP_REG_MSTAT_COUNTER_ROLLOVER_1, &val1);
pmread(mac, SUNI1x10GEXP_REG_MSTAT_COUNTER_ROLLOVER_2, &val2);
pmread(mac, SUNI1x10GEXP_REG_MSTAT_COUNTER_ROLLOVER_3, &val3);
ro = ((u64)val0 & 0xffff) | (((u64)val1 & 0xffff) << 16) |
(((u64)val2 & 0xffff) << 32) | (((u64)val3 & 0xffff) << 48);
/* Rx stats */
RMON_UPDATE(mac, RxOctetsReceivedOK, RxOctetsOK);
RMON_UPDATE(mac, RxUnicastFramesReceivedOK, RxUnicastFramesOK);
RMON_UPDATE(mac, RxMulticastFramesReceivedOK, RxMulticastFramesOK);
RMON_UPDATE(mac, RxBroadcastFramesReceivedOK, RxBroadcastFramesOK);
RMON_UPDATE(mac, RxPAUSEMACCtrlFramesReceived, RxPauseFrames);
RMON_UPDATE(mac, RxFrameCheckSequenceErrors, RxFCSErrors);
RMON_UPDATE(mac, RxFramesLostDueToInternalMACErrors,
RxInternalMACRcvError);
RMON_UPDATE(mac, RxSymbolErrors, RxSymbolErrors);
RMON_UPDATE(mac, RxInRangeLengthErrors, RxInRangeLengthErrors);
RMON_UPDATE(mac, RxFramesTooLongErrors , RxFrameTooLongErrors);
RMON_UPDATE(mac, RxJabbers, RxJabberErrors);
RMON_UPDATE(mac, RxFragments, RxRuntErrors);
RMON_UPDATE(mac, RxUndersizedFrames, RxRuntErrors);
RMON_UPDATE(mac, RxJumboFramesReceivedOK, RxJumboFramesOK);
RMON_UPDATE(mac, RxJumboOctetsReceivedOK, RxJumboOctetsOK);
/* Tx stats */
RMON_UPDATE(mac, TxOctetsTransmittedOK, TxOctetsOK);
RMON_UPDATE(mac, TxFramesLostDueToInternalMACTransmissionError,
TxInternalMACXmitError);
RMON_UPDATE(mac, TxTransmitSystemError, TxFCSErrors);
RMON_UPDATE(mac, TxUnicastFramesTransmittedOK, TxUnicastFramesOK);
RMON_UPDATE(mac, TxMulticastFramesTransmittedOK, TxMulticastFramesOK);
RMON_UPDATE(mac, TxBroadcastFramesTransmittedOK, TxBroadcastFramesOK);
RMON_UPDATE(mac, TxPAUSEMACCtrlFramesTransmitted, TxPauseFrames);
RMON_UPDATE(mac, TxJumboFramesReceivedOK, TxJumboFramesOK);
RMON_UPDATE(mac, TxJumboOctetsReceivedOK, TxJumboOctetsOK);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
return &mac->stats;
}
static int pm3393_macaddress_get(struct cmac *cmac, u8 mac_addr[6])
{
memcpy(mac_addr, cmac->instance->mac_addr, 6);
return 0;
}
static int pm3393_macaddress_set(struct cmac *cmac, u8 ma[6])
{
u32 val, lo, mid, hi, enabled = cmac->instance->enabled;
/*
* MAC addr: 00:07:43:00:13:09
*
* ma[5] = 0x09
* ma[4] = 0x13
* ma[3] = 0x00
* ma[2] = 0x43
* ma[1] = 0x07
* ma[0] = 0x00
*
* The PM3393 requires byte swapping and reverse order entry
* when programming MAC addresses:
*
* low_bits[15:0] = ma[1]:ma[0]
* mid_bits[31:16] = ma[3]:ma[2]
* high_bits[47:32] = ma[5]:ma[4]
*/
/* Store local copy */
memcpy(cmac->instance->mac_addr, ma, 6);
lo = ((u32) ma[1] << 8) | (u32) ma[0];
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
mid = ((u32) ma[3] << 8) | (u32) ma[2];
hi = ((u32) ma[5] << 8) | (u32) ma[4];
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
/* Disable Rx/Tx MAC before configuring it. */
if (enabled)
pm3393_disable(cmac, MAC_DIRECTION_RX | MAC_DIRECTION_TX);
/* Set RXXG Station Address */
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_SA_15_0, lo);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_SA_31_16, mid);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_SA_47_32, hi);
/* Set TXXG Station Address */
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_SA_15_0, lo);
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_SA_31_16, mid);
pmwrite(cmac, SUNI1x10GEXP_REG_TXXG_SA_47_32, hi);
/* Setup Exact Match Filter 1 with our MAC address
*
* Must disable exact match filter before configuring it.
*/
pmread(cmac, SUNI1x10GEXP_REG_RXXG_ADDRESS_FILTER_CONTROL_0, &val);
val &= 0xff0f;
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_ADDRESS_FILTER_CONTROL_0, val);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_EXACT_MATCH_ADDR_1_LOW, lo);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_EXACT_MATCH_ADDR_1_MID, mid);
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_EXACT_MATCH_ADDR_1_HIGH, hi);
val |= 0x0090;
pmwrite(cmac, SUNI1x10GEXP_REG_RXXG_ADDRESS_FILTER_CONTROL_0, val);
if (enabled)
pm3393_enable(cmac, enabled);
return 0;
}
static void pm3393_destroy(struct cmac *cmac)
{
kfree(cmac);
}
static struct cmac_ops pm3393_ops = {
.destroy = pm3393_destroy,
.reset = pm3393_reset,
.interrupt_enable = pm3393_interrupt_enable,
.interrupt_disable = pm3393_interrupt_disable,
.interrupt_clear = pm3393_interrupt_clear,
.interrupt_handler = pm3393_interrupt_handler,
.enable = pm3393_enable_port,
.disable = pm3393_disable,
.loopback_enable = pm3393_loopback_enable,
.loopback_disable = pm3393_loopback_disable,
.set_mtu = pm3393_set_mtu,
.set_rx_mode = pm3393_set_rx_mode,
.get_speed_duplex_fc = pm3393_get_speed_duplex_fc,
.set_speed_duplex_fc = pm3393_set_speed_duplex_fc,
.statistics_update = pm3393_update_statistics,
.macaddress_get = pm3393_macaddress_get,
.macaddress_set = pm3393_macaddress_set
};
static struct cmac *pm3393_mac_create(adapter_t *adapter, int index)
{
struct cmac *cmac;
cmac = kzalloc(sizeof(*cmac) + sizeof(cmac_instance), GFP_KERNEL);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
if (!cmac)
return NULL;
cmac->ops = &pm3393_ops;
cmac->instance = (cmac_instance *) (cmac + 1);
cmac->adapter = adapter;
cmac->instance->fc = PAUSE_TX | PAUSE_RX;
t1_tpi_write(adapter, OFFSET(0x0001), 0x00008000);
t1_tpi_write(adapter, OFFSET(0x0001), 0x00000000);
t1_tpi_write(adapter, OFFSET(0x2308), 0x00009800);
t1_tpi_write(adapter, OFFSET(0x2305), 0x00001001); /* PL4IO Enable */
t1_tpi_write(adapter, OFFSET(0x2320), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2321), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2322), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2323), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2324), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2325), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2326), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2327), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2328), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x2329), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x232a), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x232b), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x232c), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x232d), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x232e), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x232f), 0x00008800);
t1_tpi_write(adapter, OFFSET(0x230d), 0x00009c00);
t1_tpi_write(adapter, OFFSET(0x2304), 0x00000202); /* PL4IO Calendar Repetitions */
t1_tpi_write(adapter, OFFSET(0x3200), 0x00008080); /* EFLX Enable */
t1_tpi_write(adapter, OFFSET(0x3210), 0x00000000); /* EFLX Channel Deprovision */
t1_tpi_write(adapter, OFFSET(0x3203), 0x00000000); /* EFLX Low Limit */
t1_tpi_write(adapter, OFFSET(0x3204), 0x00000040); /* EFLX High Limit */
t1_tpi_write(adapter, OFFSET(0x3205), 0x000002cc); /* EFLX Almost Full */
t1_tpi_write(adapter, OFFSET(0x3206), 0x00000199); /* EFLX Almost Empty */
t1_tpi_write(adapter, OFFSET(0x3207), 0x00000240); /* EFLX Cut Through Threshold */
t1_tpi_write(adapter, OFFSET(0x3202), 0x00000000); /* EFLX Indirect Register Update */
t1_tpi_write(adapter, OFFSET(0x3210), 0x00000001); /* EFLX Channel Provision */
t1_tpi_write(adapter, OFFSET(0x3208), 0x0000ffff); /* EFLX Undocumented */
t1_tpi_write(adapter, OFFSET(0x320a), 0x0000ffff); /* EFLX Undocumented */
t1_tpi_write(adapter, OFFSET(0x320c), 0x0000ffff); /* EFLX enable overflow interrupt The other bit are undocumented */
t1_tpi_write(adapter, OFFSET(0x320e), 0x0000ffff); /* EFLX Undocumented */
t1_tpi_write(adapter, OFFSET(0x2200), 0x0000c000); /* IFLX Configuration - enable */
t1_tpi_write(adapter, OFFSET(0x2201), 0x00000000); /* IFLX Channel Deprovision */
t1_tpi_write(adapter, OFFSET(0x220e), 0x00000000); /* IFLX Low Limit */
t1_tpi_write(adapter, OFFSET(0x220f), 0x00000100); /* IFLX High Limit */
t1_tpi_write(adapter, OFFSET(0x2210), 0x00000c00); /* IFLX Almost Full Limit */
t1_tpi_write(adapter, OFFSET(0x2211), 0x00000599); /* IFLX Almost Empty Limit */
t1_tpi_write(adapter, OFFSET(0x220d), 0x00000000); /* IFLX Indirect Register Update */
t1_tpi_write(adapter, OFFSET(0x2201), 0x00000001); /* IFLX Channel Provision */
t1_tpi_write(adapter, OFFSET(0x2203), 0x0000ffff); /* IFLX Undocumented */
t1_tpi_write(adapter, OFFSET(0x2205), 0x0000ffff); /* IFLX Undocumented */
t1_tpi_write(adapter, OFFSET(0x2209), 0x0000ffff); /* IFLX Enable overflow interrupt. The other bit are undocumented */
t1_tpi_write(adapter, OFFSET(0x2241), 0xfffffffe); /* PL4MOS Undocumented */
t1_tpi_write(adapter, OFFSET(0x2242), 0x0000ffff); /* PL4MOS Undocumented */
t1_tpi_write(adapter, OFFSET(0x2243), 0x00000008); /* PL4MOS Starving Burst Size */
t1_tpi_write(adapter, OFFSET(0x2244), 0x00000008); /* PL4MOS Hungry Burst Size */
t1_tpi_write(adapter, OFFSET(0x2245), 0x00000008); /* PL4MOS Transfer Size */
t1_tpi_write(adapter, OFFSET(0x2240), 0x00000005); /* PL4MOS Disable */
t1_tpi_write(adapter, OFFSET(0x2280), 0x00002103); /* PL4ODP Training Repeat and SOP rule */
t1_tpi_write(adapter, OFFSET(0x2284), 0x00000000); /* PL4ODP MAX_T setting */
t1_tpi_write(adapter, OFFSET(0x3280), 0x00000087); /* PL4IDU Enable data forward, port state machine. Set ALLOW_NON_ZERO_OLB */
t1_tpi_write(adapter, OFFSET(0x3282), 0x0000001f); /* PL4IDU Enable Dip4 check error interrupts */
t1_tpi_write(adapter, OFFSET(0x3040), 0x0c32); /* # TXXG Config */
/* For T1 use timer based Mac flow control. */
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
t1_tpi_write(adapter, OFFSET(0x304d), 0x8000);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
t1_tpi_write(adapter, OFFSET(0x2040), 0x059c); /* # RXXG Config */
Update Chelsio gige net driver. - Use extern prefix for functions required. - Removed a lot of wrappers, including t1_read/write_reg_4. - Removed various macros, using native kernel calls now. - Enumerated various #defines. - Removed a lot of shared code which is not currently used in "NIC only" mode. - Removed dead code. Documentation/networking/cxgb.txt: - Updated release notes for version 2.1.1 drivers/net/chelsio/ch_ethtool.h - removed file, no longer using ETHTOOL namespace. drivers/net/chelsio/common.h - moved code from osdep.h to common.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cphy.h - removed dead code. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/cxgb2.c - use DMA_{32,64}BIT_MASK in include/linux/dma-mapping.h. - removed unused code. - use printk message for link info resembling drivers/net/mii.c. - no longer using the MODULE_xxx namespace. - no longer using "pci_" namespace. - no longer using ETHTOOL namespace. drivers/net/chelsio/cxgb2.h - removed file, merged into common.h drivers/net/chelsio/elmer0.h - removed dead code. - added various enums. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/espi.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. drivers/net/chelsio/espi.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/gmac.h - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/mv88x201x.c - changes to sync with Chelsio TOT. drivers/net/chelsio/osdep.h - removed file, consolidation. osdep was used to translate wrapper functions since our code supports multiple OSs. removed wrappers. drivers/net/chelsio/pm3393.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/regs.h - added a few register entries for future and current feature support. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/sge.c - rewrote large portion of scatter-gather engine to stabilize performance. - using u8/u16/u32 kernel types instead of __u8/__u16/__u32 compiler types. drivers/net/chelsio/sge.h - rewrote large portion of scatter-gather engine to stabilize performance. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/subr.c - merged tp.c into subr.c - removed various macros, using native kernel calls now. - removed a lot of wrappers, including t1_read/write_reg_4. - removed unused code. drivers/net/chelsio/suni1x10gexp_regs.h - modified copyright and authorship of file. - added comment to #endif indicating which symbol it closes. drivers/net/chelsio/tp.c - removed file, merged into subr.c. drivers/net/chelsio/tp.h - removed file. include/linux/pci_ids.h - patched to include PCI_VENDOR_ID_CHELSIO 0x1425, removed define from our code.
2005-06-23 12:40:19 +07:00
t1_tpi_write(adapter, OFFSET(0x2049), 0x0001); /* # RXXG Cut Through */
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
t1_tpi_write(adapter, OFFSET(0x2070), 0x0000); /* # Disable promiscuous mode */
/* Setup Exact Match Filter 0 to allow broadcast packets.
*/
t1_tpi_write(adapter, OFFSET(0x206e), 0x0000); /* # Disable Match Enable bit */
t1_tpi_write(adapter, OFFSET(0x204a), 0xffff); /* # low addr */
t1_tpi_write(adapter, OFFSET(0x204b), 0xffff); /* # mid addr */
t1_tpi_write(adapter, OFFSET(0x204c), 0xffff); /* # high addr */
t1_tpi_write(adapter, OFFSET(0x206e), 0x0009); /* # Enable Match Enable bit */
t1_tpi_write(adapter, OFFSET(0x0003), 0x0000); /* # NO SOP/ PAD_EN setup */
t1_tpi_write(adapter, OFFSET(0x0100), 0x0ff0); /* # RXEQB disabled */
t1_tpi_write(adapter, OFFSET(0x0101), 0x0f0f); /* # No Preemphasis */
return cmac;
}
static int pm3393_mac_reset(adapter_t * adapter)
{
u32 val;
u32 x;
u32 is_pl4_reset_finished;
u32 is_pl4_outof_lock;
u32 is_xaui_mabc_pll_locked;
u32 successful_reset;
int i;
/* The following steps are required to properly reset
* the PM3393. This information is provided in the
* PM3393 datasheet (Issue 2: November 2002)
* section 13.1 -- Device Reset.
*
* The PM3393 has three types of components that are
* individually reset:
*
* DRESETB - Digital circuitry
* PL4_ARESETB - PL4 analog circuitry
* XAUI_ARESETB - XAUI bus analog circuitry
*
* Steps to reset PM3393 using RSTB pin:
*
* 1. Assert RSTB pin low ( write 0 )
* 2. Wait at least 1ms to initiate a complete initialization of device.
* 3. Wait until all external clocks and REFSEL are stable.
* 4. Wait minimum of 1ms. (after external clocks and REFEL are stable)
* 5. De-assert RSTB ( write 1 )
* 6. Wait until internal timers to expires after ~14ms.
* - Allows analog clock synthesizer(PL4CSU) to stabilize to
* selected reference frequency before allowing the digital
* portion of the device to operate.
* 7. Wait at least 200us for XAUI interface to stabilize.
* 8. Verify the PM3393 came out of reset successfully.
* Set successful reset flag if everything worked else try again
* a few more times.
*/
successful_reset = 0;
for (i = 0; i < 3 && !successful_reset; i++) {
/* 1 */
t1_tpi_read(adapter, A_ELMER0_GPO, &val);
val &= ~1;
t1_tpi_write(adapter, A_ELMER0_GPO, val);
/* 2 */
msleep(1);
/* 3 */
msleep(1);
/* 4 */
msleep(2 /*1 extra ms for safety */ );
/* 5 */
val |= 1;
t1_tpi_write(adapter, A_ELMER0_GPO, val);
/* 6 */
msleep(15 /*1 extra ms for safety */ );
/* 7 */
msleep(1);
/* 8 */
/* Has PL4 analog block come out of reset correctly? */
t1_tpi_read(adapter, OFFSET(SUNI1x10GEXP_REG_DEVICE_STATUS), &val);
is_pl4_reset_finished = (val & SUNI1x10GEXP_BITMSK_TOP_EXPIRED);
/* TBD XXX SUNI1x10GEXP_BITMSK_TOP_PL4_IS_DOOL gets locked later in the init sequence
* figure out why? */
/* Have all PL4 block clocks locked? */
x = (SUNI1x10GEXP_BITMSK_TOP_PL4_ID_DOOL
/*| SUNI1x10GEXP_BITMSK_TOP_PL4_IS_DOOL */ |
SUNI1x10GEXP_BITMSK_TOP_PL4_ID_ROOL |
SUNI1x10GEXP_BITMSK_TOP_PL4_IS_ROOL |
SUNI1x10GEXP_BITMSK_TOP_PL4_OUT_ROOL);
is_pl4_outof_lock = (val & x);
/* ??? If this fails, might be able to software reset the XAUI part
* and try to recover... thus saving us from doing another HW reset */
/* Has the XAUI MABC PLL circuitry stablized? */
is_xaui_mabc_pll_locked =
(val & SUNI1x10GEXP_BITMSK_TOP_SXRA_EXPIRED);
successful_reset = (is_pl4_reset_finished && !is_pl4_outof_lock
&& is_xaui_mabc_pll_locked);
if (netif_msg_hw(adapter))
dev_dbg(&adapter->pdev->dev,
"PM3393 HW reset %d: pl4_reset 0x%x, val 0x%x, "
"is_pl4_outof_lock 0x%x, xaui_locked 0x%x\n",
i, is_pl4_reset_finished, val,
is_pl4_outof_lock, is_xaui_mabc_pll_locked);
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
}
return successful_reset ? 0 : 1;
}
const struct gmac t1_pm3393_ops = {
.stats_update_period = STATS_TICK_SECS,
.create = pm3393_mac_create,
.reset = pm3393_mac_reset,
[PATCH] A new 10GB Ethernet Driver by Chelsio Communications A Linux driver for the Chelsio 10Gb Ethernet Network Controller by Chelsio (http://www.chelsio.com). This driver supports the Chelsio N210 NIC and is backward compatible with the Chelsio N110 model 10Gb NICs. It supports AMD64, EM64T and x86 systems. Signed-off-by: Tina Yang <tinay@chelsio.com> Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Adrian said: - my3126.c is unused (because t1_my3126_ops isn't used anywhere) - what are the EXTRA_CFLAGS in drivers/net/chelsio/Makefile for? - $(cxgb-y) in drivers/net/chelsio/Makefile seems to be unneeded - completely unused global functions: - espi.c: t1_espi_get_intr_counts - sge.c: t1_sge_get_intr_counts - the following functions can be made static: - sge.c: t1_espi_workaround - sge.c: t1_sge_tx - subr.c: __t1_tpi_read - subr.c: __t1_tpi_write - subr.c: t1_wait_op_done shemminger said: The performance recommendations in cxgb.txt are common to all fast devices, and should be in one file rather than just for this device. I would rather see ip-sysctl.txt updated or a new file on tuning recommendations started. Some of them have consequences that aren't documented well. For example, turning off TCP timestamps risks data corruption from sequence wrap. A new driver shouldn't need so may #ifdef's unless you want to putit on older vendor versions of 2.4 Some accessor and wrapper functions like: t1_pci_read_config_4 adapter_name t1_malloc are just annoying noise. Why have useless dead code like: /* Interrupt handler */ +static int pm3393_interrupt_handler(struct cmac *cmac) +{ + u32 master_intr_status; +/* + 1. Read master interrupt register. + 2. Read BLOCK's interrupt status registers. + 3. Handle BLOCK interrupts. +*/ Jeff said: step 1: kill all the OS wrappers. And do you really need hooks for multiple MACs, when only one MAC is really supported? Typically these hooks are at a higher level anyway -- struct net_device. From: Christoph Lameter <christoph@lameter Driver modified as suggested by Pekka Enberg, Stephen Hemminger and Andrian Bunk. Reduces the size of the driver to ~260k. - clean up tabs - removed my3126.c - removed 85% of suni1x10gexp_regs.h - removed 80% of regs.h - removed various calls, renamed variables/functions. - removed system specific and other wrappers (usleep, msleep) - removed dead code - dropped redundant casts in osdep.h - dropped redundant check of kfree - dropped weird code (MODVERSIONS stuff) - reduced number of #ifdefs - use kcalloc now instead of kmalloc - Add information about known issues with the driver - Add information about authors Signed-off-by: Scott Bardone <sbardone@chelsio.com> Signed-off-by: Christoph Lameter <christoph@lameter.com> Signed-off-by: Andrew Morton <akpm@osdl.org> diff -puN /dev/null Documentation/networking/cxgb.txt
2005-03-31 04:34:31 +07:00
};