GitLab

Signed-off-by: David S. Miller <davem@davemloft.net>

From: <akepner@sgi.com>

Returning from tg3_poll() without flushing the PIO write which reenables
interrupts can result in lower cpu utilization and higher throughput.  So
use a memory barrier, mmiowb(), instead of flushing the write with a PIO
read.
Signed-off-by: Arthur Kepner <akepner@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

From: Jesse Barnes <jbarnes@engr.sgi.com>

This patch originally from Greg Banks.  Some parts of the tg3 driver depend
on PIO writes arriving in order.  This patch ensures that in two key places
using the new mmiowb macro.  This not only prevents bugs (the queues can be
corrupted), but is much faster than ensuring ordering using PIO reads
(which involve a few round trips to the target bus on some platforms).

Arthur has another patch that uses mmiowb in tg3 that he posted earlier as
well.
Signed-off-by: Greg Banks <gnb@sgi.com>
Signed-off-by: Jesse Barnes <jbarnes@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Randy Dunlap <rddunlap@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Several data structures should have been marked static because
the are local to this driver.
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

use netdev_priv
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

get rid of deprecated use of MODULE_PARM
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>

This is the second (and hopefully final) iteration of the interface
change we talked about a while ago. The patch applies cleanly against
2.6.9-rc2-mm4.

This removes the second argument (buffer for storing PCI state) from
pci_{save,restore}_state since pci_dev contains such a buffer now.
Fixed all callers.

Three drivers used to pass a buffer of 256 bytes, one only 48(!). The
rest was correct. Changes were compile tested, except for Alpha.
Signed-off-by: Roger Luethi <rl@hellgate.ch>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>

Based upon a report from Matthias Merz.
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Mike Waychison <michael.waychison@sun.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

Noticed by Eric Lemoine.
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: David S. Miller <davem@davemloft.net>

People can resurrect from the revision history if
we ever want to investigate using this again and
it kills tests from several fast paths.
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Need to clear one bit at a time, so if we are
clearing both 625_CORE_CLOCK and ALTCLOCK
we first clear the latter then the final
write will clear the former.
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: David S. Miller <davem@davemloft.net>

Otherwise we do not handle properly the case
where the switch/hub does not support auto-
negotiation.  This is what was breaking 5704
hw fiber autoneg.
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: David S. Miller <davem@redhat.com>

Signed-off-by: David S. Miller <davem@redhat.com>

Signed-off-by: David S. Miller <davem@redhat.com>

Signed-off-by: David S. Miller <davem@redhat.com>

Signed-off-by: David S. Miller <davem@redhat.com>

This broke MTU changing, among other things.
Signed-off-by: David S. Miller <davem@redhat.com>

- Clearly seperate hw-assisted vs. by-hand handling.
- Seperate PHY ID vs. Serdes state.

Should fix 5704 Fibre brokenness introduced by Sun
patches.
Signed-off-by: David S. Miller <davem@redhat.com>

During receive processing, the tg3 card updates rx_producer (the
hardware position in the receive ring) and the opaque cookie. Due to PCI
rules the stores happen in order. However the cpu may reorder the reads.

In these sort of cases there is usually a data dependency between
reading the index and looking up the data (since we use the index to
load the data). I think all cpus except alpha guarantee the reads
happen in order in this case.

However in this particular case we load hw_idx, compare it to sw_idx and
then use sw_idx to locate the opaque cookie. There is no data dependency
in this case. We need a read memory barrier between the read of
rx_producer and the opaque cookie to enforce ordering.

Thanks to Olof Johansson, Michael Chan and Broadcom for their assistance
in finding this bug.
Signed-off-by: Anton Blanchard <anton@samba.org>

- Support HW autoneg on 5704.
- On serdes, no MII reg ioctl support.

Based upon a patch from Sun.