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>
It hasn't "summed" anything in over 7 years, and it's
just a straight mempcy ala skb_copy_to_linear_data()
so just get rid of it.
Signed-off-by: David S. Miller <davem@davemloft.net>
In 7d12e780e0 David Howells performed
this evolution:
"IRQ: Maintain regs pointer globally rather than passing to IRQ handlers"
He correctly updated many of the function definitions that were using this
extra regs pointer parameter but forgot to update some caller sites of
those functions. The reason the modifications was not properly done on all
drivers is that some drivers were rarely compiled because they are for
AMIGA, or that some code sites were inside #ifdefs where the option is not
set or inside #if 0.
Here is the semantic patch that found the occurences
and fixed the problem.
@ rule1 @
identifier fn;
identifier irq, dev_id;
typedef irqreturn_t;
@@
static irqreturn_t fn(int irq, void *dev_id)
{
...
}
@@
identifier rule1.fn;
expression E1, E2, E3;
@@
fn(E1, E2
- ,E3
)
Signed-off-by: Yoann Padioleau <padator@wanadoo.fr>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
One less thing for drivers writers to worry about.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Maintain a per-CPU global "struct pt_regs *" variable which can be used instead
of passing regs around manually through all ~1800 interrupt handlers in the
Linux kernel.
The regs pointer is used in few places, but it potentially costs both stack
space and code to pass it around. On the FRV arch, removing the regs parameter
from all the genirq function results in a 20% speed up of the IRQ exit path
(ie: from leaving timer_interrupt() to leaving do_IRQ()).
Where appropriate, an arch may override the generic storage facility and do
something different with the variable. On FRV, for instance, the address is
maintained in GR28 at all times inside the kernel as part of general exception
handling.
Having looked over the code, it appears that the parameter may be handed down
through up to twenty or so layers of functions. Consider a USB character
device attached to a USB hub, attached to a USB controller that posts its
interrupts through a cascaded auxiliary interrupt controller. A character
device driver may want to pass regs to the sysrq handler through the input
layer which adds another few layers of parameter passing.
I've build this code with allyesconfig for x86_64 and i386. I've runtested the
main part of the code on FRV and i386, though I can't test most of the drivers.
I've also done partial conversion for powerpc and MIPS - these at least compile
with minimal configurations.
This will affect all archs. Mostly the changes should be relatively easy.
Take do_IRQ(), store the regs pointer at the beginning, saving the old one:
struct pt_regs *old_regs = set_irq_regs(regs);
And put the old one back at the end:
set_irq_regs(old_regs);
Don't pass regs through to generic_handle_irq() or __do_IRQ().
In timer_interrupt(), this sort of change will be necessary:
- update_process_times(user_mode(regs));
- profile_tick(CPU_PROFILING, regs);
+ update_process_times(user_mode(get_irq_regs()));
+ profile_tick(CPU_PROFILING);
I'd like to move update_process_times()'s use of get_irq_regs() into itself,
except that i386, alone of the archs, uses something other than user_mode().
Some notes on the interrupt handling in the drivers:
(*) input_dev() is now gone entirely. The regs pointer is no longer stored in
the input_dev struct.
(*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does
something different depending on whether it's been supplied with a regs
pointer or not.
(*) Various IRQ handler function pointers have been moved to type
irq_handler_t.
Signed-Off-By: David Howells <dhowells@redhat.com>
(cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
gcc4 doesn't like us declaring a static function inside another
function. We can do away with this construct altogether and use
BUILD_BUG_ON() instead (idea from Andi Kleen.)
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
Patch from Lennert Buytenhek
Move the uengine loader from arch/arm/mach-ixp2000 to arch/arm/common
so that ixp23xx can use it too.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Sky2 update changed name of netif_schedule_test to __netif_schedule_prep
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
After initialising, report the MAC address that we're using for
each port.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
The sky2 driver introduced netif_rx_schedule_test(). This is
exactly what we need, so remove our local version of this function
(which was called netif_rx_schedule_prep_notup) and use the generic
one instead.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
Do not register our netdevices with the kernel until we've actually
finished setting up the hardware and microcode.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
The way the hardware and firmware work is that there is one shared RX
queue and IRQ for a number of different network interfaces. Due to this,
we would like to process received packets for every interface in the same
NAPI poll handler, so we need a pseudo-device to schedule polling on.
What the driver currently does is that it always schedules polling for
the first network interface in the list, and processes packets for every
interface in the poll handler for that first interface -- however, this
scheme breaks down if the first network interface happens to not be up,
since netif_rx_schedule_prep() checks netif_running().
sky2 apparently has the same issue, and Stephen Hemminger suggested a
way to work around this: create a variant of netif_rx_schedule_prep()
that does not check netif_running(). I implemented this locally and
called it netif_rx_schedule_prep_notup(), and it seems to work well,
but it's something that probably not everyone would be happy with.
The ixp2000 is an ARM CPU with a high-speed network interface in the
CPU itself (full duplex 4Gb/s or 10Gb/s depending on the IXP model.)
The CPU package also contains 8 or 16 (again depending on the IXP
model) 'microengines', which are somewhat primitive but very fast
and efficient processor cores which can be used to offload various
things from the main CPU.
This driver makes the high-speed network interface in the CPU visible
and usable as a regular linux network device. Currently, it only
supports the Radisys ENP2611 IXP board, but adding support for other
board types should be fairly easy.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
