s/functin/function/ in copy pasted comment.

There is already a fast-path in pmap_extract() for the kernel pmap, so
don't bother doing a conditional for kernel vs. user-space here.

KASSERT() that pmap_extract() succeeds; it is a programming error if
it does not, and it's not a great idea to insert a garbage address
into the SGMAP page table.

Liberally sprinkle static around to get more symbols out of the
global namespace. A small bit of const poisoning in the TC code.

branches: 1.71.4;
malloc(9) -> kmem(9)

branches: 1.70.2;
Add some bus_dma instrumentation.

move common/bus_dma/ -> dev/bus_dma/

branches: 1.68.2; 1.68.12;
#include <sys/bus.h> instead of <machine/bus.h>.

Use bus_dmamem common.

vm_page: put TAILQ_ENTRY into a union with LIST_ENTRY, so we can use both.

branches: 1.65.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.64.40; 1.64.42; 1.64.44;
fix fallout from caddr_t changes.

Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.62.20;
merge yamt-uio_vmspace branch.

- use vmspace rather than proc or lwp where appropriate.
the latter is more natural to specify an address space.
(and less likely to be abused for random purposes.)
- fix a swdmover race.

branches: 1.61.2; 1.61.4; 1.61.6;
merge ktrace-lwp.

bus_dmamem_map: honour BUS_DMA_NOWAIT. noted by Manuel Bouyer.
bus_space_map: always do NOWAIT allocation as it used to be before yamt-km.

we have too many copies!

branches: 1.59.6;
Check for stale flags in the DMA map. This was causing crashes on an ES40
with more than 1GB of memory due to a bug in one of the drivers. From
Jason Thorpe.

branches: 1.58.2;
merge yamt-km branch.
- don't use managed mappings/backing objects for wired memory allocations.
save some resources like pv_entry. also fix (most of) PR/27030.
- simplify kernel memory management API.
- simplify pmap bootstrap of some ports.
- some related cleanups.

branches: 1.57.2;
Add a dm_maxsegsz public member to bus_dmamap_t. This allows a user of the API
to select the maximum segment size for each bus_dmamap_load (up to the maxsegsz
supplied to bus_dmamap_create). dm_maxsegsz is reset to the value supplied to
bus_dmamap_create when the dmamap is unloaded.

Add an extra `i' to notifes/notifed.

branches: 1.55.2; 1.55.10; 1.55.12;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.

Bump the kernel rev up to 1.6V

Add support for using cached mbuf and cluster phys addrs.

Use PAGE_SIZE rather than NBPG.

In _bus_dmamem_alloc_range(), respect the high end of the caller's range.

move initialization of the "struct pglist" returned by uvm_pglistalloc()
from the calling code into uvm_pglistalloc() itself for consistency
and easier error handling

branches: 1.49.2;
Keep track of which DMA window was actually used to map the
request (not always the passed in DMA tag if we try direct-map
and then fall back to sgmap-mapped). Use the actual window
when performing dmamap_sync and dmamap_unload operations.

Fixes DMA resource leak on systems with 2G+ RAM. Thanks to
Matt Thomas for help debugging this.

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.47.2; 1.47.4;
bzero -> memset

branches: 1.46.2;
replace vm_page_t with struct vm_page *.

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

s/vm_page_alloc_memory/uvm_pglistalloc/ in panic message.

branches: 1.43.2;
Some more slight cleanup.

Fix a couple of comments.

ANSI'ify.

When computing the segment size, consider the maximum segment
size in the DMA map.

remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

Change the kernel mmap interface so that the offset to map is an
"off_t" and the return value is a "paddr_t" to allow mappings
at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which
only changed the offset to a "vm_offset_t".

Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.

branches: 1.36.4;
Implement an internal DMA map flag which prevents coalescing of
adjacent DMA segments.

fix misspelling

Slight changes so that bounce buffers can be implemented for the Jensen.

Update for pmap_enter() API change.

branches: 1.32.2; 1.32.4; 1.32.8;
eliminate the PMAP_NEW option by making it required for all ports.
ports which previously had no support for PMAP_NEW now implement
the pmap_k* interfaces as wrappers around the non-k versions.

Change the pmap_extract() interface to:
boolean_t pmap_extract(pmap_t, vaddr_t, paddr_t *);
This makes it possible for the pmap to map physical address 0.

bus_dmamem_map() maps DMA safe memory, which is usually one or more
managed pages, into KVA space. Since the pages are managed, we should
use pmap_enter(), not pmap_kenter_pa().

Also, when entering the mappings, enter with an access_type of
VM_PROT_READ | VM_PROT_WRITE. We do this for a couple of reasons:

(1) On systems that have H/W mod/ref attributes, the hardware
may not be able to track mod/ref done by a bus master.

(2) On systems that have to do mod/ref emulation, this prevents
a mod/ref page fault from potentially happening while in an
interrupt context, which can be problematic.

This latter change is fairly important if we ever want to be able to
transfer DMA-safe memory pages to anonymous memory objects; we will need
to know that the pages are modified, or else data could be lost!

Note that while the pages are unowned (i.e. "just DMA-safe memory pages"),
they won't consume any swap resources, as the mappings are wired, and
the pages aren't on the active or inactive queues.

branches: 1.29.2; 1.29.4;
Changes for modified pmap_enter() API:
* Map the message buffer with access_type = VM_PROT_READ|VM_PROT_WRITE `just
because'.
* Map the file system buffers with access_type = VM_PROT_READ|VM_PROT_WRITE to
avoid possible problems with pagemove().
* Do not use VM_PROT_EXEC with either of the above.
* Map pages for /dev/mem with access_type = prot. Also, DO NOT use
pmap_kenter() for this, as we DO NOT want to lose modification information.
* Map pages in dumpsys() with VM_PROT_READ.
* Map pages in m68k mappedcopyin()/mappedcopyout() and writeback() with
access_type = prot.
* For now, bus_dma*(), pmap_map(), vmapbuf(), and similar functions still use
access_type = 0. This should probably be revisited.

completely remove Mach VM support. all that is left is the all the
header files as UVM still uses (most of) these.

Make bus_dmamap_load_uio() work, from Kevin Lahey <kml@nas.nasa.gov>.

Some internal cleanup; pass a tag around, not individual arguments.

vm_offset_t -> {paddr_t,vaddr_t}, vm_size_t -> vsize_t

Implement bus_dmamap_load_uio().

Allow the DMA tag to specify a boundary contraint. If the device has a
more strict boundary, the map will use it, otherwise the map will inherit
the tag's, unless the tag's constraint is 0 (no boundary constraint).

Sync the bus_dmamap_load back-end functions between alpha and i386. (only
slight, unnecessary differences)

more optimizations

Add boundary checks to _bus_dma_load...

Add support for chaining DMA windows together, for falling back on
SGMAPs if a direct-mapped window fails.

Simplify the direct-mapped DMA case somewhat by adding a window base
member to the DMA tag, and calling the direct-mapped back-ends directly,
rather than through chipset-specific front-ends which pass the window
base as an additional argument.

Use pmap_kenter_pa() in _bus_dmamem_map() if PMAP_NEW.

Add support for UVM.

Use kernel_map in bus_dmamem_{map,free}().

Use M_DMAMAP where appropriate.

Flush the CPU write buffer in _bus_dmamap_sync().

Add offset and length parameters to bus_dmamap_sync(), used for specifiying
partial syncs of a DMA mapping.

Change the last argument of bus_dmamap_sync() from bus_dmasync_op_t to int,
and allow more than one synchronization operation to be specified in
a single call. Dissallow mixing of PRE and POST operations.

Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC toBUS_DMA_COHERENT.

Implement bus_dmamap_load_mbuf().

If mapping only one segment of DMA memory into KVA space, use K0SEG.

Use avail_start/avail_end.

Implement bus_dmamem_mmap().

Remove references to vm_pmap.

Nuke the idea of <machine/options.h>. It completely defeats the purpose
of fine-grain option dependencies.

Protect access to kmem_map with splimp()/splx(). From Chuck Cranor.

branches: 1.2.2; 1.2.6;
Pull thorpej-bus-dma branch into mainline.

branches: 1.1.2;
file bus_dma.c was initially added on branch thorpej-bus-dma.

There is already a fast-path in pmap_extract() for the kernel pmap, so
don't bother doing a conditional for kernel vs. user-space here.

KASSERT() that pmap_extract() succeeds; it is a programming error if
it does not, and it's not a great idea to insert a garbage address
into the SGMAP page table.

Liberally sprinkle static around to get more symbols out of the
global namespace. A small bit of const poisoning in the TC code.

branches: 1.71.4;
malloc(9) -> kmem(9)

branches: 1.70.2;
Add some bus_dma instrumentation.

move common/bus_dma/ -> dev/bus_dma/

branches: 1.68.2; 1.68.12;
#include <sys/bus.h> instead of <machine/bus.h>.

Use bus_dmamem common.

vm_page: put TAILQ_ENTRY into a union with LIST_ENTRY, so we can use both.

branches: 1.65.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.64.40; 1.64.42; 1.64.44;
fix fallout from caddr_t changes.

Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.62.20;
merge yamt-uio_vmspace branch.

- use vmspace rather than proc or lwp where appropriate.
the latter is more natural to specify an address space.
(and less likely to be abused for random purposes.)
- fix a swdmover race.

branches: 1.61.2; 1.61.4; 1.61.6;
merge ktrace-lwp.

bus_dmamem_map: honour BUS_DMA_NOWAIT. noted by Manuel Bouyer.
bus_space_map: always do NOWAIT allocation as it used to be before yamt-km.

we have too many copies!

branches: 1.59.6;
Check for stale flags in the DMA map. This was causing crashes on an ES40
with more than 1GB of memory due to a bug in one of the drivers. From
Jason Thorpe.

branches: 1.58.2;
merge yamt-km branch.
- don't use managed mappings/backing objects for wired memory allocations.
save some resources like pv_entry. also fix (most of) PR/27030.
- simplify kernel memory management API.
- simplify pmap bootstrap of some ports.
- some related cleanups.

branches: 1.57.2;
Add a dm_maxsegsz public member to bus_dmamap_t. This allows a user of the API
to select the maximum segment size for each bus_dmamap_load (up to the maxsegsz
supplied to bus_dmamap_create). dm_maxsegsz is reset to the value supplied to
bus_dmamap_create when the dmamap is unloaded.

Add an extra `i' to notifes/notifed.

branches: 1.55.2; 1.55.10; 1.55.12;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.

Bump the kernel rev up to 1.6V

Add support for using cached mbuf and cluster phys addrs.

Use PAGE_SIZE rather than NBPG.

In _bus_dmamem_alloc_range(), respect the high end of the caller's range.

move initialization of the "struct pglist" returned by uvm_pglistalloc()
from the calling code into uvm_pglistalloc() itself for consistency
and easier error handling

branches: 1.49.2;
Keep track of which DMA window was actually used to map the
request (not always the passed in DMA tag if we try direct-map
and then fall back to sgmap-mapped). Use the actual window
when performing dmamap_sync and dmamap_unload operations.

Fixes DMA resource leak on systems with 2G+ RAM. Thanks to
Matt Thomas for help debugging this.

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.47.2; 1.47.4;
bzero -> memset

branches: 1.46.2;
replace vm_page_t with struct vm_page *.

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

s/vm_page_alloc_memory/uvm_pglistalloc/ in panic message.

branches: 1.43.2;
Some more slight cleanup.

Fix a couple of comments.

ANSI'ify.

When computing the segment size, consider the maximum segment
size in the DMA map.

remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

Change the kernel mmap interface so that the offset to map is an
"off_t" and the return value is a "paddr_t" to allow mappings
at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which
only changed the offset to a "vm_offset_t".

Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.

branches: 1.36.4;
Implement an internal DMA map flag which prevents coalescing of
adjacent DMA segments.

fix misspelling

Slight changes so that bounce buffers can be implemented for the Jensen.

Update for pmap_enter() API change.

branches: 1.32.2; 1.32.4; 1.32.8;
eliminate the PMAP_NEW option by making it required for all ports.
ports which previously had no support for PMAP_NEW now implement
the pmap_k* interfaces as wrappers around the non-k versions.

Change the pmap_extract() interface to:
boolean_t pmap_extract(pmap_t, vaddr_t, paddr_t *);
This makes it possible for the pmap to map physical address 0.

bus_dmamem_map() maps DMA safe memory, which is usually one or more
managed pages, into KVA space. Since the pages are managed, we should
use pmap_enter(), not pmap_kenter_pa().

Also, when entering the mappings, enter with an access_type of
VM_PROT_READ | VM_PROT_WRITE. We do this for a couple of reasons:

(1) On systems that have H/W mod/ref attributes, the hardware
may not be able to track mod/ref done by a bus master.

(2) On systems that have to do mod/ref emulation, this prevents
a mod/ref page fault from potentially happening while in an
interrupt context, which can be problematic.

This latter change is fairly important if we ever want to be able to
transfer DMA-safe memory pages to anonymous memory objects; we will need
to know that the pages are modified, or else data could be lost!

Note that while the pages are unowned (i.e. "just DMA-safe memory pages"),
they won't consume any swap resources, as the mappings are wired, and
the pages aren't on the active or inactive queues.

branches: 1.29.2; 1.29.4;
Changes for modified pmap_enter() API:
* Map the message buffer with access_type = VM_PROT_READ|VM_PROT_WRITE `just
because'.
* Map the file system buffers with access_type = VM_PROT_READ|VM_PROT_WRITE to
avoid possible problems with pagemove().
* Do not use VM_PROT_EXEC with either of the above.
* Map pages for /dev/mem with access_type = prot. Also, DO NOT use
pmap_kenter() for this, as we DO NOT want to lose modification information.
* Map pages in dumpsys() with VM_PROT_READ.
* Map pages in m68k mappedcopyin()/mappedcopyout() and writeback() with
access_type = prot.
* For now, bus_dma*(), pmap_map(), vmapbuf(), and similar functions still use
access_type = 0. This should probably be revisited.

completely remove Mach VM support. all that is left is the all the
header files as UVM still uses (most of) these.

Make bus_dmamap_load_uio() work, from Kevin Lahey <kml@nas.nasa.gov>.

Some internal cleanup; pass a tag around, not individual arguments.

vm_offset_t -> {paddr_t,vaddr_t}, vm_size_t -> vsize_t

Implement bus_dmamap_load_uio().

Allow the DMA tag to specify a boundary contraint. If the device has a
more strict boundary, the map will use it, otherwise the map will inherit
the tag's, unless the tag's constraint is 0 (no boundary constraint).

Sync the bus_dmamap_load back-end functions between alpha and i386. (only
slight, unnecessary differences)

more optimizations

Add boundary checks to _bus_dma_load...

Add support for chaining DMA windows together, for falling back on
SGMAPs if a direct-mapped window fails.

Simplify the direct-mapped DMA case somewhat by adding a window base
member to the DMA tag, and calling the direct-mapped back-ends directly,
rather than through chipset-specific front-ends which pass the window
base as an additional argument.

Use pmap_kenter_pa() in _bus_dmamem_map() if PMAP_NEW.

Add support for UVM.

Use kernel_map in bus_dmamem_{map,free}().

Use M_DMAMAP where appropriate.

Flush the CPU write buffer in _bus_dmamap_sync().

Add offset and length parameters to bus_dmamap_sync(), used for specifiying
partial syncs of a DMA mapping.

Change the last argument of bus_dmamap_sync() from bus_dmasync_op_t to int,
and allow more than one synchronization operation to be specified in
a single call. Dissallow mixing of PRE and POST operations.

Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC toBUS_DMA_COHERENT.

Implement bus_dmamap_load_mbuf().

If mapping only one segment of DMA memory into KVA space, use K0SEG.

Use avail_start/avail_end.

Implement bus_dmamem_mmap().

Remove references to vm_pmap.

Nuke the idea of <machine/options.h>. It completely defeats the purpose
of fine-grain option dependencies.

Protect access to kmem_map with splimp()/splx(). From Chuck Cranor.

branches: 1.2.2; 1.2.6;
Pull thorpej-bus-dma branch into mainline.

branches: 1.1.2;
file bus_dma.c was initially added on branch thorpej-bus-dma.

Liberally sprinkle static around to get more symbols out of the
global namespace. A small bit of const poisoning in the TC code.

malloc(9) -> kmem(9)

branches: 1.70.2;
Add some bus_dma instrumentation.

move common/bus_dma/ -> dev/bus_dma/

branches: 1.68.2; 1.68.12;
#include <sys/bus.h> instead of <machine/bus.h>.

Use bus_dmamem common.

vm_page: put TAILQ_ENTRY into a union with LIST_ENTRY, so we can use both.

branches: 1.65.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.64.40; 1.64.42; 1.64.44;
fix fallout from caddr_t changes.

Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.62.20;
merge yamt-uio_vmspace branch.

- use vmspace rather than proc or lwp where appropriate.
the latter is more natural to specify an address space.
(and less likely to be abused for random purposes.)
- fix a swdmover race.

branches: 1.61.2; 1.61.4; 1.61.6;
merge ktrace-lwp.

bus_dmamem_map: honour BUS_DMA_NOWAIT. noted by Manuel Bouyer.
bus_space_map: always do NOWAIT allocation as it used to be before yamt-km.

we have too many copies!

branches: 1.59.6;
Check for stale flags in the DMA map. This was causing crashes on an ES40
with more than 1GB of memory due to a bug in one of the drivers. From
Jason Thorpe.

branches: 1.58.2;
merge yamt-km branch.
- don't use managed mappings/backing objects for wired memory allocations.
save some resources like pv_entry. also fix (most of) PR/27030.
- simplify kernel memory management API.
- simplify pmap bootstrap of some ports.
- some related cleanups.

branches: 1.57.2;
Add a dm_maxsegsz public member to bus_dmamap_t. This allows a user of the API
to select the maximum segment size for each bus_dmamap_load (up to the maxsegsz
supplied to bus_dmamap_create). dm_maxsegsz is reset to the value supplied to
bus_dmamap_create when the dmamap is unloaded.

Add an extra `i' to notifes/notifed.

branches: 1.55.2; 1.55.10; 1.55.12;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.

Bump the kernel rev up to 1.6V

Add support for using cached mbuf and cluster phys addrs.

Use PAGE_SIZE rather than NBPG.

In _bus_dmamem_alloc_range(), respect the high end of the caller's range.

move initialization of the "struct pglist" returned by uvm_pglistalloc()
from the calling code into uvm_pglistalloc() itself for consistency
and easier error handling

branches: 1.49.2;
Keep track of which DMA window was actually used to map the
request (not always the passed in DMA tag if we try direct-map
and then fall back to sgmap-mapped). Use the actual window
when performing dmamap_sync and dmamap_unload operations.

Fixes DMA resource leak on systems with 2G+ RAM. Thanks to
Matt Thomas for help debugging this.

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.47.2; 1.47.4;
bzero -> memset

branches: 1.46.2;
replace vm_page_t with struct vm_page *.

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

s/vm_page_alloc_memory/uvm_pglistalloc/ in panic message.

branches: 1.43.2;
Some more slight cleanup.

Fix a couple of comments.

ANSI'ify.

When computing the segment size, consider the maximum segment
size in the DMA map.

remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

Change the kernel mmap interface so that the offset to map is an
"off_t" and the return value is a "paddr_t" to allow mappings
at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which
only changed the offset to a "vm_offset_t".

Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.

branches: 1.36.4;
Implement an internal DMA map flag which prevents coalescing of
adjacent DMA segments.

fix misspelling

Slight changes so that bounce buffers can be implemented for the Jensen.

Update for pmap_enter() API change.

branches: 1.32.2; 1.32.4; 1.32.8;
eliminate the PMAP_NEW option by making it required for all ports.
ports which previously had no support for PMAP_NEW now implement
the pmap_k* interfaces as wrappers around the non-k versions.

Change the pmap_extract() interface to:
boolean_t pmap_extract(pmap_t, vaddr_t, paddr_t *);
This makes it possible for the pmap to map physical address 0.

bus_dmamem_map() maps DMA safe memory, which is usually one or more
managed pages, into KVA space. Since the pages are managed, we should
use pmap_enter(), not pmap_kenter_pa().

Also, when entering the mappings, enter with an access_type of
VM_PROT_READ | VM_PROT_WRITE. We do this for a couple of reasons:

(1) On systems that have H/W mod/ref attributes, the hardware
may not be able to track mod/ref done by a bus master.

(2) On systems that have to do mod/ref emulation, this prevents
a mod/ref page fault from potentially happening while in an
interrupt context, which can be problematic.

This latter change is fairly important if we ever want to be able to
transfer DMA-safe memory pages to anonymous memory objects; we will need
to know that the pages are modified, or else data could be lost!

Note that while the pages are unowned (i.e. "just DMA-safe memory pages"),
they won't consume any swap resources, as the mappings are wired, and
the pages aren't on the active or inactive queues.

branches: 1.29.2; 1.29.4;
Changes for modified pmap_enter() API:
* Map the message buffer with access_type = VM_PROT_READ|VM_PROT_WRITE `just
because'.
* Map the file system buffers with access_type = VM_PROT_READ|VM_PROT_WRITE to
avoid possible problems with pagemove().
* Do not use VM_PROT_EXEC with either of the above.
* Map pages for /dev/mem with access_type = prot. Also, DO NOT use
pmap_kenter() for this, as we DO NOT want to lose modification information.
* Map pages in dumpsys() with VM_PROT_READ.
* Map pages in m68k mappedcopyin()/mappedcopyout() and writeback() with
access_type = prot.
* For now, bus_dma*(), pmap_map(), vmapbuf(), and similar functions still use
access_type = 0. This should probably be revisited.

completely remove Mach VM support. all that is left is the all the
header files as UVM still uses (most of) these.

Make bus_dmamap_load_uio() work, from Kevin Lahey <kml@nas.nasa.gov>.

Some internal cleanup; pass a tag around, not individual arguments.

vm_offset_t -> {paddr_t,vaddr_t}, vm_size_t -> vsize_t

Implement bus_dmamap_load_uio().

Allow the DMA tag to specify a boundary contraint. If the device has a
more strict boundary, the map will use it, otherwise the map will inherit
the tag's, unless the tag's constraint is 0 (no boundary constraint).

Sync the bus_dmamap_load back-end functions between alpha and i386. (only
slight, unnecessary differences)

more optimizations

Add boundary checks to _bus_dma_load...

Add support for chaining DMA windows together, for falling back on
SGMAPs if a direct-mapped window fails.

Simplify the direct-mapped DMA case somewhat by adding a window base
member to the DMA tag, and calling the direct-mapped back-ends directly,
rather than through chipset-specific front-ends which pass the window
base as an additional argument.

Use pmap_kenter_pa() in _bus_dmamem_map() if PMAP_NEW.

Add support for UVM.

Use kernel_map in bus_dmamem_{map,free}().

Use M_DMAMAP where appropriate.

Flush the CPU write buffer in _bus_dmamap_sync().

Add offset and length parameters to bus_dmamap_sync(), used for specifiying
partial syncs of a DMA mapping.

Change the last argument of bus_dmamap_sync() from bus_dmasync_op_t to int,
and allow more than one synchronization operation to be specified in
a single call. Dissallow mixing of PRE and POST operations.

Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC toBUS_DMA_COHERENT.

Implement bus_dmamap_load_mbuf().

If mapping only one segment of DMA memory into KVA space, use K0SEG.

Use avail_start/avail_end.

Implement bus_dmamem_mmap().

Remove references to vm_pmap.

Nuke the idea of <machine/options.h>. It completely defeats the purpose
of fine-grain option dependencies.

Protect access to kmem_map with splimp()/splx(). From Chuck Cranor.

branches: 1.2.2; 1.2.6;
Pull thorpej-bus-dma branch into mainline.

branches: 1.1.2;
file bus_dma.c was initially added on branch thorpej-bus-dma.

malloc(9) -> kmem(9)

Add some bus_dma instrumentation.

move common/bus_dma/ -> dev/bus_dma/

branches: 1.68.2; 1.68.12;
#include <sys/bus.h> instead of <machine/bus.h>.

Use bus_dmamem common.

vm_page: put TAILQ_ENTRY into a union with LIST_ENTRY, so we can use both.

branches: 1.65.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.64.40; 1.64.42; 1.64.44;
fix fallout from caddr_t changes.

Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.62.20;
merge yamt-uio_vmspace branch.

- use vmspace rather than proc or lwp where appropriate.
the latter is more natural to specify an address space.
(and less likely to be abused for random purposes.)
- fix a swdmover race.

branches: 1.61.2; 1.61.4; 1.61.6;
merge ktrace-lwp.

bus_dmamem_map: honour BUS_DMA_NOWAIT. noted by Manuel Bouyer.
bus_space_map: always do NOWAIT allocation as it used to be before yamt-km.

we have too many copies!

branches: 1.59.6;
Check for stale flags in the DMA map. This was causing crashes on an ES40
with more than 1GB of memory due to a bug in one of the drivers. From
Jason Thorpe.

branches: 1.58.2;
merge yamt-km branch.
- don't use managed mappings/backing objects for wired memory allocations.
save some resources like pv_entry. also fix (most of) PR/27030.
- simplify kernel memory management API.
- simplify pmap bootstrap of some ports.
- some related cleanups.

branches: 1.57.2;
Add a dm_maxsegsz public member to bus_dmamap_t. This allows a user of the API
to select the maximum segment size for each bus_dmamap_load (up to the maxsegsz
supplied to bus_dmamap_create). dm_maxsegsz is reset to the value supplied to
bus_dmamap_create when the dmamap is unloaded.

Add an extra `i' to notifes/notifed.

branches: 1.55.2; 1.55.10; 1.55.12;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.

Bump the kernel rev up to 1.6V

Add support for using cached mbuf and cluster phys addrs.

Use PAGE_SIZE rather than NBPG.

In _bus_dmamem_alloc_range(), respect the high end of the caller's range.

move initialization of the "struct pglist" returned by uvm_pglistalloc()
from the calling code into uvm_pglistalloc() itself for consistency
and easier error handling

branches: 1.49.2;
Keep track of which DMA window was actually used to map the
request (not always the passed in DMA tag if we try direct-map
and then fall back to sgmap-mapped). Use the actual window
when performing dmamap_sync and dmamap_unload operations.

Fixes DMA resource leak on systems with 2G+ RAM. Thanks to
Matt Thomas for help debugging this.

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.47.2; 1.47.4;
bzero -> memset

branches: 1.46.2;
replace vm_page_t with struct vm_page *.

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

s/vm_page_alloc_memory/uvm_pglistalloc/ in panic message.

branches: 1.43.2;
Some more slight cleanup.

Fix a couple of comments.

ANSI'ify.

When computing the segment size, consider the maximum segment
size in the DMA map.

remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

Change the kernel mmap interface so that the offset to map is an
"off_t" and the return value is a "paddr_t" to allow mappings
at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which
only changed the offset to a "vm_offset_t".

Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.

branches: 1.36.4;
Implement an internal DMA map flag which prevents coalescing of
adjacent DMA segments.

fix misspelling

Slight changes so that bounce buffers can be implemented for the Jensen.

Update for pmap_enter() API change.

branches: 1.32.2; 1.32.4; 1.32.8;
eliminate the PMAP_NEW option by making it required for all ports.
ports which previously had no support for PMAP_NEW now implement
the pmap_k* interfaces as wrappers around the non-k versions.

Change the pmap_extract() interface to:
boolean_t pmap_extract(pmap_t, vaddr_t, paddr_t *);
This makes it possible for the pmap to map physical address 0.

bus_dmamem_map() maps DMA safe memory, which is usually one or more
managed pages, into KVA space. Since the pages are managed, we should
use pmap_enter(), not pmap_kenter_pa().

Also, when entering the mappings, enter with an access_type of
VM_PROT_READ | VM_PROT_WRITE. We do this for a couple of reasons:

(1) On systems that have H/W mod/ref attributes, the hardware
may not be able to track mod/ref done by a bus master.

(2) On systems that have to do mod/ref emulation, this prevents
a mod/ref page fault from potentially happening while in an
interrupt context, which can be problematic.

This latter change is fairly important if we ever want to be able to
transfer DMA-safe memory pages to anonymous memory objects; we will need
to know that the pages are modified, or else data could be lost!

Note that while the pages are unowned (i.e. "just DMA-safe memory pages"),
they won't consume any swap resources, as the mappings are wired, and
the pages aren't on the active or inactive queues.

branches: 1.29.2; 1.29.4;
Changes for modified pmap_enter() API:
* Map the message buffer with access_type = VM_PROT_READ|VM_PROT_WRITE `just
because'.
* Map the file system buffers with access_type = VM_PROT_READ|VM_PROT_WRITE to
avoid possible problems with pagemove().
* Do not use VM_PROT_EXEC with either of the above.
* Map pages for /dev/mem with access_type = prot. Also, DO NOT use
pmap_kenter() for this, as we DO NOT want to lose modification information.
* Map pages in dumpsys() with VM_PROT_READ.
* Map pages in m68k mappedcopyin()/mappedcopyout() and writeback() with
access_type = prot.
* For now, bus_dma*(), pmap_map(), vmapbuf(), and similar functions still use
access_type = 0. This should probably be revisited.

completely remove Mach VM support. all that is left is the all the
header files as UVM still uses (most of) these.

Make bus_dmamap_load_uio() work, from Kevin Lahey <kml@nas.nasa.gov>.

Some internal cleanup; pass a tag around, not individual arguments.

vm_offset_t -> {paddr_t,vaddr_t}, vm_size_t -> vsize_t

Implement bus_dmamap_load_uio().

Allow the DMA tag to specify a boundary contraint. If the device has a
more strict boundary, the map will use it, otherwise the map will inherit
the tag's, unless the tag's constraint is 0 (no boundary constraint).

Sync the bus_dmamap_load back-end functions between alpha and i386. (only
slight, unnecessary differences)

more optimizations

Add boundary checks to _bus_dma_load...

Add support for chaining DMA windows together, for falling back on
SGMAPs if a direct-mapped window fails.

Simplify the direct-mapped DMA case somewhat by adding a window base
member to the DMA tag, and calling the direct-mapped back-ends directly,
rather than through chipset-specific front-ends which pass the window
base as an additional argument.

Use pmap_kenter_pa() in _bus_dmamem_map() if PMAP_NEW.

Add support for UVM.

Use kernel_map in bus_dmamem_{map,free}().

Use M_DMAMAP where appropriate.

Flush the CPU write buffer in _bus_dmamap_sync().

Add offset and length parameters to bus_dmamap_sync(), used for specifiying
partial syncs of a DMA mapping.

Change the last argument of bus_dmamap_sync() from bus_dmasync_op_t to int,
and allow more than one synchronization operation to be specified in
a single call. Dissallow mixing of PRE and POST operations.

Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC toBUS_DMA_COHERENT.

Implement bus_dmamap_load_mbuf().

If mapping only one segment of DMA memory into KVA space, use K0SEG.

Use avail_start/avail_end.

Implement bus_dmamem_mmap().

Remove references to vm_pmap.

Nuke the idea of <machine/options.h>. It completely defeats the purpose
of fine-grain option dependencies.

Protect access to kmem_map with splimp()/splx(). From Chuck Cranor.

branches: 1.2.2; 1.2.6;
Pull thorpej-bus-dma branch into mainline.

branches: 1.1.2;
file bus_dma.c was initially added on branch thorpej-bus-dma.

Add some bus_dma instrumentation.

move common/bus_dma/ -> dev/bus_dma/

branches: 1.68.2; 1.68.12;
#include <sys/bus.h> instead of <machine/bus.h>.

Use bus_dmamem common.

vm_page: put TAILQ_ENTRY into a union with LIST_ENTRY, so we can use both.

branches: 1.65.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.64.40; 1.64.42; 1.64.44;
fix fallout from caddr_t changes.

Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.62.20;
merge yamt-uio_vmspace branch.

- use vmspace rather than proc or lwp where appropriate.
the latter is more natural to specify an address space.
(and less likely to be abused for random purposes.)
- fix a swdmover race.

branches: 1.61.2; 1.61.4; 1.61.6;
merge ktrace-lwp.

bus_dmamem_map: honour BUS_DMA_NOWAIT. noted by Manuel Bouyer.
bus_space_map: always do NOWAIT allocation as it used to be before yamt-km.

we have too many copies!

branches: 1.59.6;
Check for stale flags in the DMA map. This was causing crashes on an ES40
with more than 1GB of memory due to a bug in one of the drivers. From
Jason Thorpe.

branches: 1.58.2;
merge yamt-km branch.
- don't use managed mappings/backing objects for wired memory allocations.
save some resources like pv_entry. also fix (most of) PR/27030.
- simplify kernel memory management API.
- simplify pmap bootstrap of some ports.
- some related cleanups.

branches: 1.57.2;
Add a dm_maxsegsz public member to bus_dmamap_t. This allows a user of the API
to select the maximum segment size for each bus_dmamap_load (up to the maxsegsz
supplied to bus_dmamap_create). dm_maxsegsz is reset to the value supplied to
bus_dmamap_create when the dmamap is unloaded.

Add an extra `i' to notifes/notifed.

branches: 1.55.2; 1.55.10; 1.55.12;
Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

Pass lwp pointers throughtout the kernel, as required, so that the lwpid can
be inserted into ktrace records. The general change has been to replace
"struct proc *" with "struct lwp *" in various function prototypes, pass
the lwp through and use l_proc to get the process pointer when needed.

Bump the kernel rev up to 1.6V

Add support for using cached mbuf and cluster phys addrs.

Use PAGE_SIZE rather than NBPG.

In _bus_dmamem_alloc_range(), respect the high end of the caller's range.

move initialization of the "struct pglist" returned by uvm_pglistalloc()
from the calling code into uvm_pglistalloc() itself for consistency
and easier error handling

branches: 1.49.2;
Keep track of which DMA window was actually used to map the
request (not always the passed in DMA tag if we try direct-map
and then fall back to sgmap-mapped). Use the actual window
when performing dmamap_sync and dmamap_unload operations.

Fixes DMA resource leak on systems with 2G+ RAM. Thanks to
Matt Thomas for help debugging this.

Update pmap_update to now take the updated pmap as an argument.
This will allow improvements to the pmaps so that they can more easily defer expensive operations, eg tlb/cache flush, til the last possible moment.

Currently this is a no-op on most platforms, so they should see no difference.

Reviewed by Jason.

branches: 1.47.2; 1.47.4;
bzero -> memset

branches: 1.46.2;
replace vm_page_t with struct vm_page *.

Sprinkle pmap_update() calls after calls to:
- pmap_enter()
- pmap_remove()
- pmap_protect()
- pmap_kenter_pa()
- pmap_kremove()
as described in pmap(9).

These calls are relatively conservative. It may be possible to
optimize these a little more.

s/vm_page_alloc_memory/uvm_pglistalloc/ in panic message.

branches: 1.43.2;
Some more slight cleanup.

Fix a couple of comments.

ANSI'ify.

When computing the segment size, consider the maximum segment
size in the DMA map.

remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>

remove/move more mach vm header files:

<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>

also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.

Change the kernel mmap interface so that the offset to map is an
"off_t" and the return value is a "paddr_t" to allow mappings
at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which
only changed the offset to a "vm_offset_t".

Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.

branches: 1.36.4;
Implement an internal DMA map flag which prevents coalescing of
adjacent DMA segments.

fix misspelling

Slight changes so that bounce buffers can be implemented for the Jensen.

Update for pmap_enter() API change.

branches: 1.32.2; 1.32.4; 1.32.8;
eliminate the PMAP_NEW option by making it required for all ports.
ports which previously had no support for PMAP_NEW now implement
the pmap_k* interfaces as wrappers around the non-k versions.

Change the pmap_extract() interface to:
boolean_t pmap_extract(pmap_t, vaddr_t, paddr_t *);
This makes it possible for the pmap to map physical address 0.

bus_dmamem_map() maps DMA safe memory, which is usually one or more
managed pages, into KVA space. Since the pages are managed, we should
use pmap_enter(), not pmap_kenter_pa().

Also, when entering the mappings, enter with an access_type of
VM_PROT_READ | VM_PROT_WRITE. We do this for a couple of reasons:

(1) On systems that have H/W mod/ref attributes, the hardware
may not be able to track mod/ref done by a bus master.

(2) On systems that have to do mod/ref emulation, this prevents
a mod/ref page fault from potentially happening while in an
interrupt context, which can be problematic.

This latter change is fairly important if we ever want to be able to
transfer DMA-safe memory pages to anonymous memory objects; we will need
to know that the pages are modified, or else data could be lost!

Note that while the pages are unowned (i.e. "just DMA-safe memory pages"),
they won't consume any swap resources, as the mappings are wired, and
the pages aren't on the active or inactive queues.

branches: 1.29.2; 1.29.4;
Changes for modified pmap_enter() API:
* Map the message buffer with access_type = VM_PROT_READ|VM_PROT_WRITE `just
because'.
* Map the file system buffers with access_type = VM_PROT_READ|VM_PROT_WRITE to
avoid possible problems with pagemove().
* Do not use VM_PROT_EXEC with either of the above.
* Map pages for /dev/mem with access_type = prot. Also, DO NOT use
pmap_kenter() for this, as we DO NOT want to lose modification information.
* Map pages in dumpsys() with VM_PROT_READ.
* Map pages in m68k mappedcopyin()/mappedcopyout() and writeback() with
access_type = prot.
* For now, bus_dma*(), pmap_map(), vmapbuf(), and similar functions still use
access_type = 0. This should probably be revisited.

completely remove Mach VM support. all that is left is the all the
header files as UVM still uses (most of) these.

Make bus_dmamap_load_uio() work, from Kevin Lahey <kml@nas.nasa.gov>.

Some internal cleanup; pass a tag around, not individual arguments.

vm_offset_t -> {paddr_t,vaddr_t}, vm_size_t -> vsize_t

Implement bus_dmamap_load_uio().

Allow the DMA tag to specify a boundary contraint. If the device has a
more strict boundary, the map will use it, otherwise the map will inherit
the tag's, unless the tag's constraint is 0 (no boundary constraint).

Sync the bus_dmamap_load back-end functions between alpha and i386. (only
slight, unnecessary differences)

more optimizations

Add boundary checks to _bus_dma_load...

Add support for chaining DMA windows together, for falling back on
SGMAPs if a direct-mapped window fails.

Simplify the direct-mapped DMA case somewhat by adding a window base
member to the DMA tag, and calling the direct-mapped back-ends directly,
rather than through chipset-specific front-ends which pass the window
base as an additional argument.

Use pmap_kenter_pa() in _bus_dmamem_map() if PMAP_NEW.

Add support for UVM.

Use kernel_map in bus_dmamem_{map,free}().

Use M_DMAMAP where appropriate.

Flush the CPU write buffer in _bus_dmamap_sync().

Add offset and length parameters to bus_dmamap_sync(), used for specifiying
partial syncs of a DMA mapping.

Change the last argument of bus_dmamap_sync() from bus_dmasync_op_t to int,
and allow more than one synchronization operation to be specified in
a single call. Dissallow mixing of PRE and POST operations.

Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC toBUS_DMA_COHERENT.

Implement bus_dmamap_load_mbuf().

If mapping only one segment of DMA memory into KVA space, use K0SEG.

Use avail_start/avail_end.

Implement bus_dmamem_mmap().

Remove references to vm_pmap.

Nuke the idea of <machine/options.h>. It completely defeats the purpose
of fine-grain option dependencies.

Protect access to kmem_map with splimp()/splx(). From Chuck Cranor.

branches: 1.2.2; 1.2.6;
Pull thorpej-bus-dma branch into mainline.

branches: 1.1.2;
file bus_dma.c was initially added on branch thorpej-bus-dma.