Drop the first argument of intr_establish().

Instead, require all callers to put the right value in the ih_pil field, and
have intr_establish() trust them rather than assigning this field again from
its first argument.

ok claudio@ kettenis@

send_softint() was designed as being able to target a particular cpu, but the
code for this was never written and all uses target the running cpu anyway,
so stop pretending it may do things it won't do and drop that cpu argument.

ok claudio@ kettenis@

Move C declarations from intreg.h to <machine/intr.h> which includes it.

This makes intreg.h locore-friendly - it only contains the MAXINTNUM define
after that.

ok claudio@ kettenis@

Remove unused spllock().

OK deraadt@ mpi@

Define and use IPL_MPFLOOR in our common mutex implementation.

ok kettenis@, visa@

we have some spare interrupt levels, so move softnet above softclock.

ok kettenis@

rework sparc64 splfoo functions to be more consistent with other archs

this also moves us toward having an MI splraise().

sparc64 (and sparc) are different to the other archs because they
have macros that build templates. each spl uses that macro to create
an instance of an inline function specific to that spl call.

this moves it to having a single splraise inline that the spl api is
defined with. eg, #define splfoo() _splraise(IPL_FOO).

ok kettenis@

Store the target CPU in "struct intrhand" and use it in intr_barrier().
Also use it wherever we configure the hardware to direct interrupts to the
right CPU.

Introduce intr_barrier(4), an interface that guarantees that an interrupt
handler that was running has finished.

ok miod@, guenther@

Implement IPL_MPSAFE for pci bus interrupt handlers.

Implement support for running interrupt handlers without taking the kernel
lock, by adding a new BUS_INTR_ESTABLISH_MPSAFE flag for use with the
(sparc64-specific) bus_intr_establish(9) interface. Add support for this
flag to schizo(4); other bus drivers will ignore it for now. While there,
remove the BUS_INTR_ESTABLISH_FASTTRAP flag which serves no purpose.

Including <sparc64/intreg.h> here is ugly, but intr.h only comes from
<machine/cpu.h> if _KERNEL is defined. The bootblocks are stupid, and
for now they need this. We need more namespace cleanup, for sure.

Normalize sentinel. Use _MACHINE_*_H_ and _<ARCH>_*_H_ properly and consitently.

Discussed and okay drahn@. Okay deraadt@.

First pass at removing clauses 3 and 4 from NetBSD licenses.

Not sure what's more surprising: how long it took for NetBSD to
catch up to the rest of the BSDs (including UCB), or the amount of
code that NetBSD has claimed for itself without attributing to the
actual authors.

OK deraadt@

Rename IPL_SOFTSERIAL to IPL_SOFTTTY. Sprinkle some KNF while I'm there.

Introduce a per-handler interrupt acknowledgement function.

Make sure an interrupt handler does not get on the per-cpu list of pending
interrupts twice, with one exception: interrupt handlers are allowed to be on
the tail of said lists (needed for clock interrupts on MP kernels).
Prevents losing interrupts. Makes usb keyboard as console work on Sun Blade
1000/2000 with MP kernels.

Use new event counter API for interrupt counting on sparc64. deraadt@ tholo@
drahn@ millert@ ok

tabs vs spaces

Add a "where" argument to the sparc64 interrupt code. This lets us
associate a name with each interrupt handler. This is not visible
outside the kernel (yet).

ok jason@

define IPL_STATCLOCK.

Implement splassert just like it's on sparc.
This also requires a change to rename {PIL,IPL}_IMP to {PIL,IPL}_VM.
XXX - We should get rid of PIL_*.

Final __P removal plus some cosmetic fixups

First round of __P removal in sys

branches: 1.1.1; 1.1.2; 1.1.4;
Initial revision

Remove unused spllock().

OK deraadt@ mpi@

Define and use IPL_MPFLOOR in our common mutex implementation.

ok kettenis@, visa@

we have some spare interrupt levels, so move softnet above softclock.

ok kettenis@

rework sparc64 splfoo functions to be more consistent with other archs

this also moves us toward having an MI splraise().

sparc64 (and sparc) are different to the other archs because they
have macros that build templates. each spl uses that macro to create
an instance of an inline function specific to that spl call.

this moves it to having a single splraise inline that the spl api is
defined with. eg, #define splfoo() _splraise(IPL_FOO).

ok kettenis@

Store the target CPU in "struct intrhand" and use it in intr_barrier().
Also use it wherever we configure the hardware to direct interrupts to the
right CPU.

Introduce intr_barrier(4), an interface that guarantees that an interrupt
handler that was running has finished.

ok miod@, guenther@

Implement IPL_MPSAFE for pci bus interrupt handlers.

Implement support for running interrupt handlers without taking the kernel
lock, by adding a new BUS_INTR_ESTABLISH_MPSAFE flag for use with the
(sparc64-specific) bus_intr_establish(9) interface. Add support for this
flag to schizo(4); other bus drivers will ignore it for now. While there,
remove the BUS_INTR_ESTABLISH_FASTTRAP flag which serves no purpose.

Including <sparc64/intreg.h> here is ugly, but intr.h only comes from
<machine/cpu.h> if _KERNEL is defined. The bootblocks are stupid, and
for now they need this. We need more namespace cleanup, for sure.

Normalize sentinel. Use _MACHINE_*_H_ and _<ARCH>_*_H_ properly and consitently.

Discussed and okay drahn@. Okay deraadt@.

First pass at removing clauses 3 and 4 from NetBSD licenses.

Not sure what's more surprising: how long it took for NetBSD to
catch up to the rest of the BSDs (including UCB), or the amount of
code that NetBSD has claimed for itself without attributing to the
actual authors.

OK deraadt@

Rename IPL_SOFTSERIAL to IPL_SOFTTTY. Sprinkle some KNF while I'm there.

Introduce a per-handler interrupt acknowledgement function.

Make sure an interrupt handler does not get on the per-cpu list of pending
interrupts twice, with one exception: interrupt handlers are allowed to be on
the tail of said lists (needed for clock interrupts on MP kernels).
Prevents losing interrupts. Makes usb keyboard as console work on Sun Blade
1000/2000 with MP kernels.

Use new event counter API for interrupt counting on sparc64. deraadt@ tholo@
drahn@ millert@ ok

tabs vs spaces

Add a "where" argument to the sparc64 interrupt code. This lets us
associate a name with each interrupt handler. This is not visible
outside the kernel (yet).

ok jason@

define IPL_STATCLOCK.

Implement splassert just like it's on sparc.
This also requires a change to rename {PIL,IPL}_IMP to {PIL,IPL}_VM.
XXX - We should get rid of PIL_*.

Final __P removal plus some cosmetic fixups

First round of __P removal in sys

branches: 1.1.1; 1.1.2; 1.1.4;
Initial revision

Define and use IPL_MPFLOOR in our common mutex implementation.

ok kettenis@, visa@

we have some spare interrupt levels, so move softnet above softclock.

ok kettenis@

rework sparc64 splfoo functions to be more consistent with other archs

this also moves us toward having an MI splraise().

sparc64 (and sparc) are different to the other archs because they
have macros that build templates. each spl uses that macro to create
an instance of an inline function specific to that spl call.

this moves it to having a single splraise inline that the spl api is
defined with. eg, #define splfoo() _splraise(IPL_FOO).

ok kettenis@

Store the target CPU in "struct intrhand" and use it in intr_barrier().
Also use it wherever we configure the hardware to direct interrupts to the
right CPU.

Introduce intr_barrier(4), an interface that guarantees that an interrupt
handler that was running has finished.

ok miod@, guenther@

Implement IPL_MPSAFE for pci bus interrupt handlers.

Implement support for running interrupt handlers without taking the kernel
lock, by adding a new BUS_INTR_ESTABLISH_MPSAFE flag for use with the
(sparc64-specific) bus_intr_establish(9) interface. Add support for this
flag to schizo(4); other bus drivers will ignore it for now. While there,
remove the BUS_INTR_ESTABLISH_FASTTRAP flag which serves no purpose.

Including <sparc64/intreg.h> here is ugly, but intr.h only comes from
<machine/cpu.h> if _KERNEL is defined. The bootblocks are stupid, and
for now they need this. We need more namespace cleanup, for sure.

Normalize sentinel. Use _MACHINE_*_H_ and _<ARCH>_*_H_ properly and consitently.

Discussed and okay drahn@. Okay deraadt@.

First pass at removing clauses 3 and 4 from NetBSD licenses.

Not sure what's more surprising: how long it took for NetBSD to
catch up to the rest of the BSDs (including UCB), or the amount of
code that NetBSD has claimed for itself without attributing to the
actual authors.

OK deraadt@

Rename IPL_SOFTSERIAL to IPL_SOFTTTY. Sprinkle some KNF while I'm there.

Introduce a per-handler interrupt acknowledgement function.

Make sure an interrupt handler does not get on the per-cpu list of pending
interrupts twice, with one exception: interrupt handlers are allowed to be on
the tail of said lists (needed for clock interrupts on MP kernels).
Prevents losing interrupts. Makes usb keyboard as console work on Sun Blade
1000/2000 with MP kernels.

Use new event counter API for interrupt counting on sparc64. deraadt@ tholo@
drahn@ millert@ ok

tabs vs spaces

Add a "where" argument to the sparc64 interrupt code. This lets us
associate a name with each interrupt handler. This is not visible
outside the kernel (yet).

ok jason@

define IPL_STATCLOCK.

Implement splassert just like it's on sparc.
This also requires a change to rename {PIL,IPL}_IMP to {PIL,IPL}_VM.
XXX - We should get rid of PIL_*.

Final __P removal plus some cosmetic fixups

First round of __P removal in sys

branches: 1.1.1; 1.1.2; 1.1.4;
Initial revision