Commit graph

48 commits

Author SHA1 Message Date
Stephen Hemminger
bea3348eef [NET]: Make NAPI polling independent of struct net_device objects.
Several devices have multiple independant RX queues per net
device, and some have a single interrupt doorbell for several
queues.

In either case, it's easier to support layouts like that if the
structure representing the poll is independant from the net
device itself.

The signature of the ->poll() call back goes from:

	int foo_poll(struct net_device *dev, int *budget)

to

	int foo_poll(struct napi_struct *napi, int budget)

The caller is returned the number of RX packets processed (or
the number of "NAPI credits" consumed if you want to get
abstract).  The callee no longer messes around bumping
dev->quota, *budget, etc. because that is all handled in the
caller upon return.

The napi_struct is to be embedded in the device driver private data
structures.

Furthermore, it is the driver's responsibility to disable all NAPI
instances in it's ->stop() device close handler.  Since the
napi_struct is privatized into the driver's private data structures,
only the driver knows how to get at all of the napi_struct instances
it may have per-device.

With lots of help and suggestions from Rusty Russell, Roland Dreier,
Michael Chan, Jeff Garzik, and Jamal Hadi Salim.

Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra,
Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan.

[ Ported to current tree and all drivers converted.  Integrated
  Stephen's follow-on kerneldoc additions, and restored poll_list
  handling to the old style to fix mutual exclusion issues.  -DaveM ]

Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-10 16:47:45 -07:00
Michael Chan
e30372c912 [BNX2]: Support NVRAM on 5709.
The NVRAM interface is slightly modified on the 5709.  To properly
support it, we need to change the buffered flag in the flash data
structure into multiple flags to indicate buffered operation, address
translation, and the use of write enable (WREN).  The 5709 flash
only requires the buffered operation bit to be set.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-07-18 01:46:46 -07:00
Michael Chan
58fc2ea405 [BNX2]: Print management firmware version.
Add management firmware version for ethtool -i.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-07-10 22:18:38 -07:00
Michael Chan
df149d70e1 [BNX2]: Enhance the heartbeat.
In addition to the periodic heartbeat, we're adding a heartbeat
request interrupt when the heartbeat is late.  This is needed during
netpoll where the timer is not available.  -rt kernels will also
benefit since the timer is not as accurate.

[ We discussed this patch last time and we decided that the -rt
  kernel problem alone did not justify this patch.  I think the
  netpoll problem makes this patch necessary. ]

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-07-10 22:18:36 -07:00
Michael Chan
9700e6befe [BNX2]: Add remote PHY bit definitions.
Add new fields in struct bnx2 and other bit definitions in shared
memory to support remote PHY.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-07-10 22:18:31 -07:00
Michael Chan
0aa38df7cd [BNX2]: Enable DMA on 5709.
Add missing code to enable DMA on 5709 A1.  The bit is a no-op on A0
and therefore can be set on all 5709 chips.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Acked-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-06-07 13:38:38 -07:00
Michael Chan
72fbaeb623 [BNX2]: Update version and reldate.
Update version to 1.5.10.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:25:32 -07:00
Michael Chan
883e515118 [BNX2]: Print bus information for PCIE devices.
Fix the code to print PCI or PCIE bus information for all devices.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:25:11 -07:00
Michael Chan
8e6a72c435 [BNX2]: Add 1-shot MSI handler for 5709.
The 5709 supports the one-shot MSI handler similar to some of the tg3
chips.  In this mode, the MSI disables itself automatically until it
is re-enabled at the end of NAPI poll.

Put the request_irq/free_irq logic in common procedures.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:24:48 -07:00
Michael Chan
1b8227c48e [BNX2]: Add indirect spinlock.
The indirect register access method will be used by more than one
caller in BH context (NAPI poll and timer), so a spinlock is required.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:24:05 -07:00
Michael Chan
27a005b883 [BNX2]: Add support for 5709 Serdes.
Add PCI ID and code to support the 5709 Serdes PHY.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:23:41 -07:00
Michael Chan
ca58c3af99 [BNX2]: Put MII register offsets in the bnx2 struct.
The 5709 Serdes device uses non-standard MII register offsets.  This
re-structuring will make it easier to support 5709 Serdes.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:22:52 -07:00
Michael Chan
4666f87a82 [BNX2]: Add ipv6 TSO and checksum for 5709.
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:22:28 -07:00
Michael Chan
40453c839f [BNX2]: Add 40-bit DMA workaround for 5708.
The internal PCIE-to-PCIX bridge of the 5708 has the same 40-bit DMA
limitation as some of the tg3 chips.  Set dma_mask and persistent DMA
mask to 40-bit to workaround.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-05-03 13:19:18 -07:00
Michael Chan
68c9f75a05 [BNX2]: Fix occasional NETDEV WATCHDOG on 5709.
Tweak a register setting to prevent the tx mailbox from halting.

Update version to 1.5.8.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-24 15:35:53 -07:00
Michael Chan
b659f44e4e [BNX2]: PHY workaround for 5709 A0.
5709 A0 copper devices will not link up with some link partners
without this workaround.

Update driver to 1.5.5.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-02-02 00:46:35 -08:00
Michael Chan
9052a840ff [BNX2]: Add new 5709 registers (part 2).
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-12-02 21:24:25 -08:00
Michael Chan
19cdeb794b [BNX2]: Add new 5709 registers (part 1).
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-12-02 21:24:23 -08:00
Michael Chan
af3ee519c5 [BNX2]: Re-organize firmware structures.
Re-organize the firmware handling code and declarations a bit to make
the code more compact.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-12-02 21:24:22 -08:00
Michael Chan
f8dd064ee1 [BNX2]: Add 5708S parallel detection.
Add code to parallel detect 1Gbps and 2.5Gbps link speeds.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-12-02 21:24:20 -08:00
Michael Chan
80be44348e [BNX2]: Improve SerDes handling.
1. Add support for 2.5Gbps forced speed setting.
2. Remove a long udelay() loop and change to msleep().
3. Other misc. SerDes fixes.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-12-02 21:24:18 -08:00
Jeff Garzik
6aa20a2235 drivers/net: Trim trailing whitespace
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-09-13 13:24:59 -04:00
Michael Chan
2f8af120a1 [BNX2]: Fix tx race condition.
Fix a subtle race condition between bnx2_start_xmit() and bnx2_tx_int()
similar to the one in tg3 discovered by Herbert Xu:

CPU0					CPU1
bnx2_start_xmit()
	if (tx_ring_full) {
		tx_lock
					bnx2_tx()
						if (!netif_queue_stopped)
		netif_stop_queue()
		if (!tx_ring_full)
						update_tx_ring
			netif_wake_queue()
		tx_unlock
	}

Even though tx_ring is updated before the if statement in bnx2_tx_int() in
program order, it can be re-ordered by the CPU as shown above.  This
scenario can cause the tx queue to be stopped forever if bnx2_tx_int() has
just freed up the entire tx_ring.  The possibility of this happening
should be very rare though.

The following changes are made, very much identical to the tg3 fix:

1. Add memory barrier to fix the above race condition.

2. Eliminate the private tx_lock altogether and rely solely on
netif_tx_lock.  This eliminates one spinlock in bnx2_start_xmit()
when the ring is full.

3. Because of 2, use netif_tx_lock in bnx2_tx_int() before calling
netif_wake_queue().

4. Add memory barrier to bnx2_tx_avail().

5. Add bp->tx_wake_thresh which is set to half the tx ring size.

6. Check for the full wake queue condition before getting
netif_tx_lock in tg3_tx().  This reduces the number of unnecessary
spinlocks when the tx ring is full in a steady-state condition.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-08-17 16:29:51 -07:00
Michael Chan
6c4f095eae [BNX2]: Turn off link during shutdown
Minor change in shutdown logic to effect a link down.

Update version to 1.4.43.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-29 16:58:24 -07:00
Michael Chan
c86a31f452 [BNX2]: Use CPU native page size
Use CPU native page size to determine various ring sizes. This allows
order-0 memory allocations on all systems.

Added check to limit the page size to 16K since that's the maximum rx
ring size that will be used. This will prevent using unnecessarily
large page sizes on some architectures with large page sizes.
[Suggested by David Miller]

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-17 21:30:49 -07:00
Michael Chan
fba9fe911b [BNX2]: Add firmware decompression
Add functions to decompress firmware before loading to the internal
CPUs. Compressing the firmware reduces the driver size significantly.

Added file name length sanity check in the gzip header to prevent
going past the end of buffer [suggested by DaveM].

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-17 21:30:44 -07:00
Michael Chan
cea94db9b4 [BNX2]: Add an rx drop counter
Add a counter for packets dropped by firmware.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-17 21:30:40 -07:00
Michael Chan
29b12174ac [BNX2]: Separate tx producer and consumer fields
Put the tx producer and consumer fields in separate cache lines in
the device structure, similar to the VJ net channel queue structure.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-23 01:13:43 -08:00
Michael Chan
f2a4f05216 [BNX2]: Move .h files to bnx2.c
Move all #include <> from bnx2.h to bnx2.c.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-23 01:13:12 -08:00
Michael Chan
0f31f99446 [BNX2]: Combine small mem allocations
Combine two small (56 byte and 320 byte) pci consistent memory
allocations into one allocation. Jeff Garzik suggested to store
the combined size in the bp structure for later use when freeing
the memory.

Use kzalloc() instead of kmalloc() + memset().

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-23 01:12:38 -08:00
Michael Chan
bf5295bba8 [BNX2]: Fix link change handling
Fix some link-related problems by doing a coalesce_now after link
change interrupt to flush out the transient link status.

To facilitate this, the host coalesce cmd register value is cached in
the device structure.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-23 01:11:56 -08:00
Michael Chan
0d36f37bb1 [BNX2]: include <linux/vmalloc.h>
Include <linux/vmalloc.h> so that it compiles properly on all archs.

Update version to 1.4.38.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-20 17:55:25 -08:00
Michael Chan
13daffa2f2 [BNX2]: Support larger rx ring sizes (part 1)
Increase maximum receive ring size from 255 to 1020 by supporting
up to 4 linked pages of receive descriptors. To accomodate the
higher memory usage, each physical descriptor page is allocated
separately and the software ring that keeps track of the SKBs and the
DMA addresses is allocated using vmalloc.

Some of the receive-related fields in the bp structure are re-
organized a bit for better locality of reference.

The max. was reduced to 1020 from 4080 after discussion with David
Miller.

This patch contains ring init code changes only. This next patch
contains rx data path code changes.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-20 17:49:20 -08:00
Michael Chan
206cc83ccd [BNX2]: Update version and copyright year
Update version to 1.4.31 and add 2006 copyright.

Skip the last digit when reporting the firmware version.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-01-23 16:14:05 -08:00
Michael Chan
1122db717a [BNX2]: Fix nvram sizing
Add code to correctly determine nvram size.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-01-23 16:11:42 -08:00
Michael Chan
ade2bfe7d1 [BNX2]: Fix UDP checksum verification
Fix TCP/UDP checksum verification. Use status bits in the buffer
descriptor instead of the checksum value to verify rx checksum.
Using the checksum value will be incorrect if the UDP packet has
zero in the UDP checksum field.

Firmware update required for this fix.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-01-23 16:09:51 -08:00
Michael Chan
dda1e390bf [BNX2]: Misc. fixes
Some misc. fixes for WoL, 5708 B1, and a typo '=' instead of '=='.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-01-23 16:08:14 -08:00
Michael Chan
b090ae2b59 [BNX2]: Improve handshake with firmware
Improve handshake with bootcode with the following changes:

1. Increase timeout to 100msec and use msleep instead of udelay.

2. Add more error checking for timeouts and errors. 

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-01-23 16:07:10 -08:00
Michael Chan
e29054f92d [BNX2]: Fix VLAN on ASF
Always set up the device to strip incoming VLAN tags when ASF is
enabled. ASF firmware will not parse packets correctly if VLAN tags
are not stripped.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2006-01-23 16:06:06 -08:00
Michael Chan
f4e418f7f3 [PATCH] bnx2: refine bnx2_poll
Refine bnx2_poll() logic to write back the most up-to-date status tag
when all work has been processed. This eliminates some occasional
extra interrupts when a older status tag is written even though all
work has been processed.

The idea is to read the status tag just before exiting bnx2_poll() and
then check again for any new work. If no new work is pending, the
status tag written back will not generate any extra interrupt. This
logic is similar to the changes David Miller did to tg3_poll().

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-11-05 21:00:03 -05:00
Michael Chan
e3648b3d8d [PATCH] bnx2: update firmware handshake for 5708
Dynamically determine the shared memory location where eeprom
parameters are stored instead of using a fixed location.

Add speed reporting to management firmware. This allows management
firmware to know the current speed without contending for MII
registers.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-11-05 21:00:03 -05:00
Michael Chan
371377091d [PATCH] bnx2: update nvram code for 5708
Update bnx2 nvram code with support for 5708.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-11-05 21:00:02 -05:00
Michael Chan
5b0c76ad94 [PATCH] bnx2: add 5708 support
Add 5708 copper and serdes basic support, including 2.5 Gbps support
on 5708 serdes. SPEED_2500 is also added to ethtool.h

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-11-05 21:00:02 -05:00
Michael Chan
c921e4c4db [BNX2]: Fix bug in irq handler and add prefetch
Fix bug in bnx2_interrupt() that caused an unnecessary register read.
The BNX2_PCICFG_MISC_STATUS should only be read when the status tag
has not changed.

Add prefetch of the status block in bnx2_msi() similar to tg3_msi().
The status block is not touched in bnx2_msi() and prefetching it will
speed up bnx2_poll() that will run on the same CPU that received the
MSI.

Update version.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-09-08 13:15:32 -07:00
Michael Chan
e89bbf1049 [BNX2]: remove atomics in tx
Remove atomic operations in the fast tx path. Expensive atomic
operations were used to keep track of the number of available tx
descriptors. The new code uses the difference between the consumer
and producer index to determine the number of free tx descriptors.

As suggested by Jeff Garzik, the name of the inline function is
changed to all lower case.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-08-29 16:10:38 -07:00
Michael Chan
cd339a0ed6 [BNX2]: speedup serdes linkup
This speeds up link-up time on 5706 SerDes if the link partner does
not autoneg, a rather common scenario in blade servers. Some blade
servers use IPMI for keyboard input and it's important to minimize
link disruptions.

The speedup is achieved by shortening the timer to (HZ / 3) during
the transient period right after initiating a SerDes autoneg. If
autoneg does not complete, parallel detect can be done sooner. After
the transient period is over, the timer goes back to its normal HZ
interval.

As suggested by Jeff Garzik, the timer initialization is moved to
bnx2_init_board() from bnx2_open().

An eeprom bit is also added to allow default forced SerDes speed for
even faster link-up time.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-08-29 16:10:34 -07:00
Michael Chan
afdc08b9f9 [BNX2]: Fix rtnl deadlock in bnx2_close
This fixes an rtnl deadlock problem when flush_scheduled_work() is
called from bnx2_close(). In rare cases, linkwatch_event() may be on
the workqueue from a previous close of a different device and it will
try to get the rtnl lock which is already held by dev_close().

The fix is to set a flag if we are in the reset task which is run
from the workqueue. bnx2_close() will loop until the flag is cleared.
As suggested by Jeff Garzik, the loop is changed to call msleep(1)
instead of yield() in the original patch.

flush_scheduled_work() is also moved to bnx2_remove_one() before the
netdev is freed.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-08-29 16:10:29 -07:00
Michael Chan
b6016b7673 [BNX2]: New Broadcom gigabit network driver.
A new driver bnx2 for Broadcom bcm5706 is available.

The patch also includes new 1000BASE-X advertisement bit definitions in
mii.h

Thanks to David Miller and Jeff Garzik for reviewing and their valuable
feedback.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-05-26 13:03:09 -07:00