remove the old drm sources.

branches: 1.3.10;
DRM_IOCTL_SG_ALLOC needs IOWR, not IOR!

branches: 1.2.14; 1.2.16; 1.2.18; 1.2.20; 1.2.22;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

struct device * -> device_t, no functional changes intended.

Rework scatter / gather page allocation inspired by r186295 of FreeBSD SVN:

rework drm_scatter.c which allocates scatter / gather pages for use by
ati pci gart to use bus_dma to handle the allocations. This fixes
a garbled screen issue on at least some radeons (X1400 tested). It is
also likely that this is the correct fix for PR# 119324, though that
is not confirmed yet.

Fixes similar issues on a Radeon Mobility X300, non-Mobility X800 (including
a hard-lock of the machine).

With feedback from ad@, ok mrg@

add a comment explaining DRM_NETBSD_ADDR2HANDLE/DRM_NETBSD_HANDLE2ADDR:

* This hack strips the top bit from amd64 addresses, which avoid
* udv_attach() returning NULL for "negative" offset.
* A better hack would be to encode the offset of some kernel data
* structure..

add a new DRM_HANDLE_NEEDS_MASK macro to check whether the above need to
be applied for various mapping types (_DRM_SHM and _DRM_SCATTER_GATHER.)

branches: 1.32.2; 1.32.4; 1.32.6; 1.32.10;
revert unnecessary changes to 3rd party code.

split device/softc

Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.28.2;
o Split device_t/softc for agp(4).
o agp_ali.c: remove unused 'agp' member from agp_ali_softc.
o drm: agp_find_device() returns 'void *', not a device_t.
o Use device_t, cfdata_t instead of struct pointers.
o Don't cast void pointers in assignments.

Don't use MALLOC_DECLARE / MALLOC_DEFINE if building as a module.

NOTE: this commit message accompnies the last one for sys/dev/pci/drm.

Revert condvar changes; there seem to be other locking issues
(independent of these changes insofar as I can tell), and it will
only complicate things for now to introduce more changes.

Argh!

Allow building as a module.

Use cv(9) instead of mtsleep/wakeup. Also, define drm_hz as
mstohz(333) (this seems to eliminate lockups for me using the r300
driver). Tested with i915 and radeon on amd64.

branches: 1.23.2;
Revert all the changes this time.

Revert last commit -- breaks i915.

A few steps down the yellow brick road toward bus_dma-ification:

- Add drm_dmamem_alloc/drm_dmamem_free to drm_memory.c to nicely wrap up
the bus_dma API.

- Start using the above in drm_pci.c.

- Add DRM_NETBSD_DMA_ADDR/DRM_NETBSD_DMA_VADDR macros.

Locking:

- Use IPL_NONE for all locks except the IRQ lock, which runs at IPL_VM.

- Use IPL_VM instead of IPL_TTY with pci_intr_establish() for consistency's
sake. These two changes seem to eliminate the presistent lockups I was
having (NetBSD-current/amd64 r300).

- Start getting rid of DRM_SPININIT/DRM_SPINUNINIT and DRM_SPINLOCK/
DRM_SPINUNLOCK ... these annoy me to no end--not to mention that they
locks may or may not be spinlocks! It's a linux frob, really.
We're way beyond merging any useful bsd-core code on a large scale, which
was the only good reason to keep them around.

Memory allocation:

- Change drm_memory.c so that it contains generally useful, memory
allocation functions using kmem(9) (mostly used by the drivers
themselves). However, I expect to use this more in the future
in the "bsd core". These functions always use KM_NOSLEEP.
The new drm_dmamem_alloc function has a wait argument which
takes DRM_DMA_WAIT/DRM_DMA_NOWAIT (defined as their bus_dma
counterparts), and honors this hint in its calls to kmem(9)
and bus_dma(9) functions.

- Got rid of these functions' "area" argument--it's been deprecated for
ages. Provide macros in drmP.h to deal with the os-independent code.

- Declare these functions inline -- I believe they're used enough
by the i915 and radeon drivers to justify it. Please let me know
if I am mistaken.

NOTE: With these changes, a glxgears score which was previously
~3900fps is now ~4400fps (same setup as mentioned above). I realize
that using kmem(9) could cause problems, but I can't seem to run into
any with my test setup. If anyone smells regression, please let me
know.

Use aprint instead of DRM_INFO macro to cleanup attachment messages.

For _DRM_SHM allocations, do not use the VA as drm_map handle on x86_64
since udv_attach will not accept a larger offset than signed 64-bit will
provide. Not the best fix, but good enough for now -- at least an X server
will now startup and get proper mappings.

Wedge in a smattering of updates for ati_pcigart.c and radeon support:

-- detypedef drm_ati_pcigart_info
-- add support for variable sized apertures in ati_pcigart.c
-- Add support for radeon IGP adapters (untested on NetBSD)
-- Update r300_reg.h to a much newer version.
-- Miscellaneous fixes in radeon_cp.c, radeon_state.c.

TODO: r300_cmdbuf.c (?)

Please report any regressions to me. While I try to make sure the
updates are somewhat coherent, our initial code base is old enough
such that it's quite difficult to be absolutely certain. I hope to
rebase the NetBSD-specific code on the BSD code at the head of GIT
within a month or two, so this will do for now.

branches: 1.17.2;
DRM_SPINLOCK() wants a spin mutex, it may be called from interrupt
context. Switch it to IPL_VM.

branches: 1.16.2;
Change the DRM_SPININIT macro to call mutex_init with MUTEX_DEFAULT
and IPL_NONE. MUTEX_DRIVER is no longer part of the stable API. Moreover,
IPL_TTY is now IPL_VM, and this code should not be running with such a
high ipl.

Fix bus_dmamem_free() call to provide 'segs' as the second argument.
While here, add DRM_PCI_DMAADDR macro so that we set handle->busaddr
in a portable way.
Next is the more difficult [annoying] part: use bus_dma(9) with
scatter/gather buffers.

branches: 1.14.2; 1.14.6;
membar_memory -> membar_sync

Use new style memory barriers.

__FUNCTION__ -> __func__

branches: 1.11.4;
drm unbreaking, round 1:

Fix the bus_dma(9) use in drm_pci.c and chnage struct drm_dma_handle_t
to match it. Remove member dmaaddr from drm_dma_handle_t as well,
as I don't see it used anywhere. Compile-tested only for now; please
report any problems to me. Thanks.

branches: 1.10.2;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.9.2;
Until I have a chance to further evaluate and re-work the locking here,
change from an adaptive to spin mutex @ IPL_TTY. This seems to mitigate
or eliminate a lot of locking-related issues (such as a hang when running
glxgears under AIGLX). Tested by me and Vincent <10.50@free.fr>.

In this code, we have three locks: irq, dma, and device. If anyone has
any thoughts on how to improve this situation, e.g. by using different
lock types/priorities/soft interrupts/etc. for different locks, please
share them with me.

NOTE: I've had the best results thusfar using Mesa 7.0.2-rc1 (r300 users
could even try the r300 driver code from the master branch) and xorg-server
1.4+patches from GIT.

First round of DRM improvements: switch to mutex(9), make use of &
operator consistent in locking macros, use mb(9) for memory barrier ops
(to be replaced with bus_space_barrier(9) at a later date; this is not
necessary for x86 but might be for e.g. mips).


More to come. Try it out! It's a lot more stable now just with the
mutex(9) changes.

machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.6.2; 1.6.10; 1.6.12; 1.6.14; 1.6.18;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.5.2;
caddr_t -> char *, even though it's in #ifdef __FreeBSD__ sections, just
so grep -r won't show it any more.

Free ressources more agressively.
(The natural place for this would be on each close, but we don't have
a driver entry point for this.)
from Yorick Hardy

Use void * rather than caddr_t to fix the build.

In drm_ioremap, provide extra slots for mapping AGP memory space; it's not
reasonable to assume that these regions will show up in a VGA BAR.

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.7.10;
split device/softc

branches: 1.6.2;
o Split device_t/softc for agp(4).
o agp_ali.c: remove unused 'agp' member from agp_ali_softc.
o drm: agp_find_device() returns 'void *', not a device_t.
o Use device_t, cfdata_t instead of struct pointers.
o Don't cast void pointers in assignments.

branches: 1.5.2;
Use aprint instead of DRM_INFO macro to cleanup attachment messages.

branches: 1.4.8; 1.4.10; 1.4.12;
use __KERNEL_RCSID()

branches: 1.3.2; 1.3.4; 1.3.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.2.2; 1.2.10; 1.2.12; 1.2.14; 1.2.18; 1.2.20;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.1.2;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.7.12;
Make the atomic macros compile: do not discard 'volatile' qualification
from argument `p'. Use `b' instead of `bit'.

Rework drm_atomic.h for correctness and readability:

- Use macros where appropriate.
- Remove code for other operating systems -- we don't need it.
- In atomic_read/atomic_set, cast to volatile.
- Misc. cleanup.

Oops, did not compile. Fix it.

atomic_add_32 -> atomic_and_32. For atomic_read/atomic_write, cast
to volatile uint32_t *.

branches: 1.3.2; 1.3.4;
Adapt to recent NetBSD which has an atomic MI API. With code from
bjs@ and ad@.

branches: 1.2.14; 1.2.16;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.6.10;
Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.4.8; 1.4.12; 1.4.14; 1.4.16;
use __KERNEL_RCSID()

branches: 1.3.2; 1.3.4; 1.3.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.2.2; 1.2.10; 1.2.12; 1.2.14; 1.2.18; 1.2.20;
In drm_remove_magic, free the magic entry when it is found instead of
attempting to free a NULL reference in the error case. From Yorick Hardy.

branches: 1.1.2;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

drm_addmap():
- for _DRM_CONSISTENT mappings, keep the handle.
- use DRM_HANDLE_NEEDS_MASK()

drm_mapbufs():
- use DRM_HANDLE_NEEDS_MASK()

drm_mmap():
- use DRM_HANDLE_NEEDS_MASK()
- for _DRM_SCATTER_GATHER and _DRM_SHM, use vtophys() on the
adjusted offset. XXX


this is gets radeon working on amd64 with an older PCI 9250 card.

XXX: need to excise vtophys() usage.
XXX: need to finish porting these fixes to external.

branches: 1.10.4; 1.10.10;
Revert previous, i915drm works again.

- Don't use btop, as only three (?) platforms define this macro; use
round_page(x) >> PAGE_SHIFT instead.

- Add bus_space_mmap() in drm_vm.c to give a proper cookie to the device pager;
this gets rid of one of the two #ifdef macppc conditionals.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.7.2; 1.7.4;
For _DRM_SHM allocations, do not use the VA as drm_map handle on x86_64
since udv_attach will not accept a larger offset than signed 64-bit will
provide. Not the best fix, but good enough for now -- at least an X server
will now startup and get proper mappings.

branches: 1.6.8; 1.6.10; 1.6.12;
use __KERNEL_RCSID()

branches: 1.5.2; 1.5.4; 1.5.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.4.2; 1.4.10; 1.4.12; 1.4.14; 1.4.18; 1.4.20;
drm_do_addbufs_pci: Fix the mess of malloc()s and free()s, which
do not accept NULL arguments.
Found via CID: 4439

Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.2.2;
Pass a proper vaddr to uvm_mmap.

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

bzero -> memset

branches: 1.3.12; 1.3.20; 1.3.26;
use __KERNEL_RCSID()

branches: 1.2.2; 1.2.4; 1.2.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.7.12;
NOTE: this commit message accompnies the last one for sys/dev/pci/drm.

Revert condvar changes; there seem to be other locking issues
(independent of these changes insofar as I can tell), and it will
only complicate things for now to introduce more changes.

Argh!

Use cv(9) instead of mtsleep/wakeup. Also, define drm_hz as
mstohz(333) (this seems to eliminate lockups for me using the r300
driver). Tested with i915 and radeon on amd64.

branches: 1.5.10; 1.5.12;
use __KERNEL_RCSID()

branches: 1.4.2; 1.4.4; 1.4.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.3.2;
First round of DRM improvements: switch to mutex(9), make use of &
operator consistent in locking macros, use mb(9) for memory barrier ops
(to be replaced with bus_space_barrier(9) at a later date; this is not
necessary for x86 but might be for e.g. mips).


More to come. Try it out! It's a lot more stable now just with the
mutex(9) changes.

branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.16; 1.2.20;
Try not to free NULL pointers.

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.3.12; 1.3.26;
use __KERNEL_RCSID()

branches: 1.2.2; 1.2.4; 1.2.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

struct device * -> device_t, no functional changes intended.

Rework scatter / gather page allocation inspired by r186295 of FreeBSD SVN:

rework drm_scatter.c which allocates scatter / gather pages for use by
ati pci gart to use bus_dma to handle the allocations. This fixes
a garbled screen issue on at least some radeons (X1400 tested). It is
also likely that this is the correct fix for PR# 119324, though that
is not confirmed yet.

Fixes similar issues on a Radeon Mobility X300, non-Mobility X800 (including
a hard-lock of the machine).

With feedback from ad@, ok mrg@

Move initialisation of the irq spinlock into the device handling code,
at least the Intel driver depends on it whether the IRQ is currently
enabled or not.

branches: 1.22.2;
Pass a bus_space_tag_t as first argument of bus_space_unmap, not an int.

Change major()/minor() to return 32-bit types again, called
devmajor_t/devminor_t, as proposed on tech-kern.
This avoids 64-bit arithmetics and 64-bit printf formats in parts
of the kernel where it is not really useful, and helps clarity.

Don't attempt to unload a DRM device that's in use. (Note:
Unloading doesn't work right in any case -- it doesn't clean up the
sysctl tree, among other things. This code needs Work, but at least
this prevents it crashing randomly due to autounload while X is
running.) Also, fix the dependency list for radeondrm.

contributed anonymously.

branches: 1.19.4; 1.19.6;
split device/softc

Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.16.2;
Don't call devsw_attach/detach if _MODULE is not defined

NOTE: this commit message accompnies the last one for sys/dev/pci/drm.

Revert condvar changes; there seem to be other locking issues
(independent of these changes insofar as I can tell), and it will
only complicate things for now to introduce more changes.

Argh!

Allow building as a module.

Use cv(9) instead of mtsleep/wakeup. Also, define drm_hz as
mstohz(333) (this seems to eliminate lockups for me using the r300
driver). Tested with i915 and radeon on amd64.

branches: 1.12.2;
Use aprint instead of DRM_INFO macro to cleanup attachment messages.

make device lock adaptive
(needed because pfind() is called from inside)

branches: 1.10.2; 1.10.4;
Catch up with descriptor handling changes. See kern_descrip.c revision
1.173 for details.

-map 64-bit PCI BARs
-integrate some bits from upstream to support i965
from Yorick Hardy per PM

branches: 1.8.4; 1.8.8;
use __KERNEL_RCSID()

branches: 1.7.2; 1.7.4; 1.7.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.6.2;
First round of DRM improvements: switch to mutex(9), make use of &
operator consistent in locking macros, use mb(9) for memory barrier ops
(to be replaced with bus_space_barrier(9) at a later date; this is not
necessary for x86 but might be for e.g. mips).


More to come. Try it out! It's a lot more stable now just with the
mutex(9) changes.

branches: 1.5.2; 1.5.10; 1.5.12; 1.5.14; 1.5.18;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.4.2;
Free ressources more agressively.
(The natural place for this would be on each close, but we don't have
a driver entry point for this.)
from Yorick Hardy

Newline after foodrm0 at vga0 message.

Initialize extra AGP resource slots.

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.4.12; 1.4.26;
use __KERNEL_RCSID()

branches: 1.3.2; 1.3.4; 1.3.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.16; 1.2.20; 1.2.22;
Free ressources more agressively.
(The natural place for this would be on each close, but we don't have
a driver entry point for this.)
from Yorick Hardy

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.3.12; 1.3.26;
use __KERNEL_RCSID()

branches: 1.2.2; 1.2.4; 1.2.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

Move initialisation of the irq spinlock into the device handling code,
at least the Intel driver depends on it whether the IRQ is currently
enabled or not.

branches: 1.16.2; 1.16.4; 1.16.6; 1.16.10;
revert unnecessary changes to 3rd party code.

split device/softc

Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.12.2;
NOTE: this commit message accompnies the last one for sys/dev/pci/drm.

Revert condvar changes; there seem to be other locking issues
(independent of these changes insofar as I can tell), and it will
only complicate things for now to introduce more changes.

Argh!

Use cv(9) instead of mtsleep/wakeup. Also, define drm_hz as
mstohz(333) (this seems to eliminate lockups for me using the r300
driver). Tested with i915 and radeon on amd64.

branches: 1.10.2;
Revert last commit -- breaks i915.

A few steps down the yellow brick road toward bus_dma-ification:

- Add drm_dmamem_alloc/drm_dmamem_free to drm_memory.c to nicely wrap up
the bus_dma API.

- Start using the above in drm_pci.c.

- Add DRM_NETBSD_DMA_ADDR/DRM_NETBSD_DMA_VADDR macros.

Locking:

- Use IPL_NONE for all locks except the IRQ lock, which runs at IPL_VM.

- Use IPL_VM instead of IPL_TTY with pci_intr_establish() for consistency's
sake. These two changes seem to eliminate the presistent lockups I was
having (NetBSD-current/amd64 r300).

- Start getting rid of DRM_SPININIT/DRM_SPINUNINIT and DRM_SPINLOCK/
DRM_SPINUNLOCK ... these annoy me to no end--not to mention that they
locks may or may not be spinlocks! It's a linux frob, really.
We're way beyond merging any useful bsd-core code on a large scale, which
was the only good reason to keep them around.

Memory allocation:

- Change drm_memory.c so that it contains generally useful, memory
allocation functions using kmem(9) (mostly used by the drivers
themselves). However, I expect to use this more in the future
in the "bsd core". These functions always use KM_NOSLEEP.
The new drm_dmamem_alloc function has a wait argument which
takes DRM_DMA_WAIT/DRM_DMA_NOWAIT (defined as their bus_dma
counterparts), and honors this hint in its calls to kmem(9)
and bus_dma(9) functions.

- Got rid of these functions' "area" argument--it's been deprecated for
ages. Provide macros in drmP.h to deal with the os-independent code.

- Declare these functions inline -- I believe they're used enough
by the i915 and radeon drivers to justify it. Please let me know
if I am mistaken.

NOTE: With these changes, a glxgears score which was previously
~3900fps is now ~4400fps (same setup as mentioned above). I realize
that using kmem(9) could cause problems, but I can't seem to run into
any with my test setup. If anyone smells regression, please let me
know.

branches: 1.8.2; 1.8.4;
use aprint_*_dev and device_xname

branches: 1.7.6;
__FUNCTION__ -> __func__

use __KERNEL_RCSID()

branches: 1.5.2; 1.5.4; 1.5.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.4.2;
First round of DRM improvements: switch to mutex(9), make use of &
operator consistent in locking macros, use mb(9) for memory barrier ops
(to be replaced with bus_space_barrier(9) at a later date; this is not
necessary for x86 but might be for e.g. mips).


More to come. Try it out! It's a lot more stable now just with the
mutex(9) changes.

branches: 1.3.2; 1.3.10; 1.3.12; 1.3.14; 1.3.18;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.2.2;
remove a debug printf which I had left in accidentally,
print the standard "dv_xname: interrupting at ..." line instead

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.3.8;
Fix handling of lists with a single entry in list_for_each_safe();
this also plugs a memory leak.

branches: 1.2.14; 1.2.18; 1.2.24;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.7.12;
NOTE: this commit message accompnies the last one for sys/dev/pci/drm.

Revert condvar changes; there seem to be other locking issues
(independent of these changes insofar as I can tell), and it will
only complicate things for now to introduce more changes.

Argh!

Use cv(9) instead of mtsleep/wakeup. Also, define drm_hz as
mstohz(333) (this seems to eliminate lockups for me using the r300
driver). Tested with i915 and radeon on amd64.

branches: 1.5.10; 1.5.12;
use __KERNEL_RCSID()

branches: 1.4.2; 1.4.4;
Bring in some modest fixes from GIT that seemed appropriate to me and did not
break functionality. I've tested this on a PCIE R423 (X800).

- In drm_drv.c, check that the requested context lock matches the holder
of the lock.

- Unify radeon offset checking. (r300_cmdbuf.c, radeon_drv.h, radeon_state.c)
[Replace r300_check_offset() with generic radeon_check_offset(), which doesn't
reject valid offsets when the framebuffer area is at the very end of the card's
32 bit address space. Make radeon_check_and_fixup_offset() use
radeon_check_offset() as well.
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=7697]

NOTE: There is another AGP fix that I didn't have time to merge
that I thought might help with getting this to work on macppc.
Contact me if you're interested. Hi macallan@! ;)

radeon_cp.c, radeon_drv.h: (GIT id: bb5f2158dbd30dbbffa3881fac75b71d71ecaaf9)
- set the address to access the aperture on the CPU side correctly
[This code relied on the CPU and GPU address for the aperture being the same,
On some r5xx hardware I was playing with I noticed that this isn't always true.
I wonder if this will fix some of those r4xx DRI issues we've seen in the past.]

- Commit the ring after earch partial texture. (radeon_state.c)
(GIT ID: ac8406420ea80ffe5ccaadc1ff0124f95709a23d)
[Commit the ring after each partial texture upload blit.
This makes sure each blit starts as early as possible, which may improve
texture upload performance in some cases.]

If anyone's having any issues in particular--especially those which
can be _teased apart from AGP driver issues--please let me know
and I'll investigate. Thanks. There are also tons of fixes for the i915,
but I don't want to change too much at once.

branches: 1.3.2;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.2.2;
First round of DRM improvements: switch to mutex(9), make use of &
operator consistent in locking macros, use mb(9) for memory barrier ops
(to be replaced with bus_space_barrier(9) at a later date; this is not
necessary for x86 but might be for e.g. mips).


More to come. Try it out! It's a lot more stable now just with the
mutex(9) changes.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

Only pull in genfb.h if _KERNEL_OPT is defined. Reported by Kurt Schreiner

Use genfb_borrow; drm works with genfb now.

branches: 1.14.4; 1.14.10;
Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.12.2;
Don't use MALLOC_DECLARE / MALLOC_DEFINE if building as a module.

Allow building as a module.

branches: 1.10.2;
Revert last commit -- breaks i915.

A few steps down the yellow brick road toward bus_dma-ification:

- Add drm_dmamem_alloc/drm_dmamem_free to drm_memory.c to nicely wrap up
the bus_dma API.

- Start using the above in drm_pci.c.

- Add DRM_NETBSD_DMA_ADDR/DRM_NETBSD_DMA_VADDR macros.

Locking:

- Use IPL_NONE for all locks except the IRQ lock, which runs at IPL_VM.

- Use IPL_VM instead of IPL_TTY with pci_intr_establish() for consistency's
sake. These two changes seem to eliminate the presistent lockups I was
having (NetBSD-current/amd64 r300).

- Start getting rid of DRM_SPININIT/DRM_SPINUNINIT and DRM_SPINLOCK/
DRM_SPINUNLOCK ... these annoy me to no end--not to mention that they
locks may or may not be spinlocks! It's a linux frob, really.
We're way beyond merging any useful bsd-core code on a large scale, which
was the only good reason to keep them around.

Memory allocation:

- Change drm_memory.c so that it contains generally useful, memory
allocation functions using kmem(9) (mostly used by the drivers
themselves). However, I expect to use this more in the future
in the "bsd core". These functions always use KM_NOSLEEP.
The new drm_dmamem_alloc function has a wait argument which
takes DRM_DMA_WAIT/DRM_DMA_NOWAIT (defined as their bus_dma
counterparts), and honors this hint in its calls to kmem(9)
and bus_dma(9) functions.

- Got rid of these functions' "area" argument--it's been deprecated for
ages. Provide macros in drmP.h to deal with the os-independent code.

- Declare these functions inline -- I believe they're used enough
by the i915 and radeon drivers to justify it. Please let me know
if I am mistaken.

NOTE: With these changes, a glxgears score which was previously
~3900fps is now ~4400fps (same setup as mentioned above). I realize
that using kmem(9) could cause problems, but I can't seem to run into
any with my test setup. If anyone smells regression, please let me
know.

branches: 1.8.2; 1.8.4;
Don't include i810 hack if DRM_NO_AGP is defined, as it results in
an undefined symbol (agp_i810_borrow) at kernel link time.

-map 64-bit PCI BARs
-integrate some bits from upstream to support i965
from Yorick Hardy per PM

Add a horrible hack to allow the DRM driver to get at the bus_space_handle
allocated by the AGP driver in the case of Intel chipset graphics.
This is different from the patch by Yorick Hardy circulated a while ago
in that it doesn't change the semantics of reference counting within
the (3rd-party) DRM code -- it just bypasses it.
Needless to say that it is uglier, but it eases future updates to
the DRM code because the change there is just 2 LOC.

Imo, a cleaner solution for all that would be to attach "agp" to "vga"
in the intel chipset graphics case, which better reflects how the hardware
is structured. This would still need a hack to the DRM code, but
it would be confined to childs of the "vga" device, without need for
global variables.
Since there is a variety of intel chipsets with AGP and/or builtin
graphics options, this would need a considerable testing effort.

branches: 1.5.4; 1.5.8;
use __KERNEL_RCSID()

branches: 1.4.2; 1.4.4; 1.4.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.3.2; 1.3.10; 1.3.12; 1.3.14; 1.3.18; 1.3.20;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.2.2;
In drm_ioremap, provide extra slots for mapping AGP memory space; it's not
reasonable to assume that these regions will show up in a VGA BAR.

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

Revert previous, i915drm works again.

branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8;
Use symbolic names for mem->phase, make error reporting more sensible.

Add work-in-progress of NetBSD-specific code. I will be committing a lot
more changes shortly (as soon as I work out the mmap offset insanity).

Revert previous, i915drm works again.

branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8;
Use symbolic names for mem->phase, make error reporting more sensible.

Add work-in-progress of NetBSD-specific code. I will be committing a lot
more changes shortly (as soon as I work out the mmap offset insanity).

remove the old drm sources.

include the size in a falled allocation message.

branches: 1.13.4; 1.13.10;
Revert previous, i915drm works again.

- Don't use btop, as only three (?) platforms define this macro; use
round_page(x) >> PAGE_SHIFT instead.

- Add bus_space_mmap() in drm_vm.c to give a proper cookie to the device pager;
this gets rid of one of the two #ifdef macppc conditionals.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.10.2; 1.10.4;
Revert last commit -- breaks i915.

A few steps down the yellow brick road toward bus_dma-ification:

- Add drm_dmamem_alloc/drm_dmamem_free to drm_memory.c to nicely wrap up
the bus_dma API.

- Start using the above in drm_pci.c.

- Add DRM_NETBSD_DMA_ADDR/DRM_NETBSD_DMA_VADDR macros.

Locking:

- Use IPL_NONE for all locks except the IRQ lock, which runs at IPL_VM.

- Use IPL_VM instead of IPL_TTY with pci_intr_establish() for consistency's
sake. These two changes seem to eliminate the presistent lockups I was
having (NetBSD-current/amd64 r300).

- Start getting rid of DRM_SPININIT/DRM_SPINUNINIT and DRM_SPINLOCK/
DRM_SPINUNLOCK ... these annoy me to no end--not to mention that they
locks may or may not be spinlocks! It's a linux frob, really.
We're way beyond merging any useful bsd-core code on a large scale, which
was the only good reason to keep them around.

Memory allocation:

- Change drm_memory.c so that it contains generally useful, memory
allocation functions using kmem(9) (mostly used by the drivers
themselves). However, I expect to use this more in the future
in the "bsd core". These functions always use KM_NOSLEEP.
The new drm_dmamem_alloc function has a wait argument which
takes DRM_DMA_WAIT/DRM_DMA_NOWAIT (defined as their bus_dma
counterparts), and honors this hint in its calls to kmem(9)
and bus_dma(9) functions.

- Got rid of these functions' "area" argument--it's been deprecated for
ages. Provide macros in drmP.h to deal with the os-independent code.

- Declare these functions inline -- I believe they're used enough
by the i915 and radeon drivers to justify it. Please let me know
if I am mistaken.

NOTE: With these changes, a glxgears score which was previously
~3900fps is now ~4400fps (same setup as mentioned above). I realize
that using kmem(9) could cause problems, but I can't seem to run into
any with my test setup. If anyone smells regression, please let me
know.

branches: 1.8.4;
Fix bus_dmamem_free() call to provide 'segs' as the second argument.
While here, add DRM_PCI_DMAADDR macro so that we set handle->busaddr
in a portable way.
Next is the more difficult [annoying] part: use bus_dma(9) with
scatter/gather buffers.

branches: 1.7.4; 1.7.8;
use __KERNEL_RCSID()

branches: 1.6.2; 1.6.4;
drm unbreaking, round 1:

Fix the bus_dma(9) use in drm_pci.c and chnage struct drm_dma_handle_t
to match it. Remove member dmaaddr from drm_dma_handle_t as well,
as I don't see it used anywhere. Compile-tested only for now; please
report any problems to me. Thanks.

branches: 1.5.2;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.4.2; 1.4.4; 1.4.12; 1.4.14; 1.4.16; 1.4.20; 1.4.22;
since no boundary restrictions apply, a zero boundary argument to
bus_dmamem_alloc() is the right thing

Fix previous, use %zd.

- pass a sane boundary parameter to bus_dmamem_alloc
- sprinkle aprint_error()

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.2.18; 1.2.32;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

Rework scatter / gather page allocation inspired by r186295 of FreeBSD SVN:

rework drm_scatter.c which allocates scatter / gather pages for use by
ati pci gart to use bus_dma to handle the allocations. This fixes
a garbled screen issue on at least some radeons (X1400 tested). It is
also likely that this is the correct fix for PR# 119324, though that
is not confirmed yet.

Fixes similar issues on a Radeon Mobility X300, non-Mobility X800 (including
a hard-lock of the machine).

With feedback from ad@, ok mrg@

branches: 1.7.2; 1.7.4; 1.7.6; 1.7.10;
revert unnecessary changes to 3rd party code.

Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.4.6; 1.4.10; 1.4.12; 1.4.14;
__FUNCTION__ -> __func__

use __KERNEL_RCSID()

branches: 1.2.2; 1.2.4; 1.2.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

branches: 1.3.12; 1.3.26;
use __KERNEL_RCSID()

branches: 1.2.2; 1.2.4; 1.2.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.16; 1.1.20; 1.1.22;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove the old drm sources.

drm_addmap():
- for _DRM_CONSISTENT mappings, keep the handle.
- use DRM_HANDLE_NEEDS_MASK()

drm_mapbufs():
- use DRM_HANDLE_NEEDS_MASK()

drm_mmap():
- use DRM_HANDLE_NEEDS_MASK()
- for _DRM_SCATTER_GATHER and _DRM_SHM, use vtophys() on the
adjusted offset. XXX


this is gets radeon working on amd64 with an older PCI 9250 card.

XXX: need to excise vtophys() usage.
XXX: need to finish porting these fixes to external.

move a DRM_SPINUNLOCK() to where it will actually be executed.

branches: 1.15.4; 1.15.10;
Revert previous, i915drm works again.

Revert to using atop(offset + map->offset) for map types _DRM_AGP,
_DRM_REGISTERS, and _DRM_FRAME_BUFFER. Fixes issue where a call
to DrmMap() in libdrm returns EINVAL.

make an i386 ALL kernel build again

- Don't use btop, as only three (?) platforms define this macro; use
round_page(x) >> PAGE_SHIFT instead.

- Add bus_space_mmap() in drm_vm.c to give a proper cookie to the device pager;
this gets rid of one of the two #ifdef macppc conditionals.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.10.2; 1.10.4;
Ugh, unbreak _DRM_SHM on amd64 again.

Fix compilation on i386.

For _DRM_SHM allocations, do not use the VA as drm_map handle on x86_64
since udv_attach will not accept a larger offset than signed 64-bit will
provide. Not the best fix, but good enough for now -- at least an X server
will now startup and get proper mappings.

Returning EINVAL from drm_mmap isn't a smart thing to do.

branches: 1.6.8; 1.6.10; 1.6.12;
use __KERNEL_RCSID()

branches: 1.5.2; 1.5.4; 1.5.6;
Remove #ifdef __OpSys__ sections to make this readable. Add RCS tags.
Nothing done in header files, nor in dev/pci/drm (we might not want to
touch this for the sake of merging) yet. Compile tested; please let
me know if this broke anything.

branches: 1.4.2; 1.4.10; 1.4.12; 1.4.14; 1.4.18; 1.4.20;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.3.2;
revert previous - it doesn't work on i386 at least

we return addresses, not page numbers in *mmap() methods

Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

add VirtualBox DRM driver

branches: 1.8.4; 1.8.6; 1.8.8;
remove the old drm sources.

you can now easily select trying external drm with "options DRM_EXTERNAL".

branches: 1.6.4; 1.6.8; 1.6.12;
support new drm.

XXX: not fully enabled yet. config(1) has a makeoptions bug
with complex conditionals. see files.drm / files.pcidrm for
some instructions on how to enable it.

remove some more local changes of the form "foo.h" -> <dev/*/foo.h>
by using makeoptions in files.drm/files.pcidrm. this reduces the
diffs against other drm code by about 40kb.

branches: 1.4.2;
Revert previous, i915drm works again.

- Use the extent manager to track mappings and provide sane offset
"cookies" for mmap. Now the radeon driver works on amd64!

- Use bus_dma for nearly all map types (carried over from prior patch).

- Plug some lock leaks in drm_bufs.c and remove superfluous locking in
drm_add_magic() (from git).

- Lower spl of mutexes to IPL_NONE (and so are now adaptive instead of
spin mutexes), save for irq_lock which is a spin mutex at IPL_VM.
Use mutex_spin_enter()/mutex_spin_exit() for irq_lock.

ok jmcneill@

branches: 1.2.2; 1.2.14; 1.2.34; 1.2.38; 1.2.40; 1.2.42;
Introduce options DRM_NO_AGP and DRM_NO_MTRR, so I can run this on
platforms that don't provide these services.

In drm_mmap, don't wrap the return value with atop() if we're on macppc.

While we're here, fix:
pci_intr_establish(&dev->pa.pa_pc, ...);
to:
pci_intr_establish(dev->pa.pa_pc, ...);

The former doesn't even compile on macppc, and I'm amazed that it works at
all anywhere else.

branches: 1.1.2;
Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11.
Minor modifications by me:
-use an mi device major number
-(coarsly) divided into pci card specific and less specific parts, moved
the latter to dev/drm
-renamed autoconf attributes to reflect this
Todo:
-adapt all card frontends but i915 to drm include file location
-review the mtrr change
-make the change to agp_i810.c coexist with the fix for buggy VESA
BIOSes which is commented out temporarily
-RCS IDs etc style stuff
-LKM support (rescan support for vga)
-test

remove checks for failure after memory allocation calls that cannot fail:

kmem_alloc() with KM_SLEEP
kmem_zalloc() with KM_SLEEP
percpu_alloc()
pserialize_create()
psref_class_create()

all of these paths include an assertion that the allocation has not failed,
so callers should not assert that again.

Check pmf_device_register return value. NFC
Appeases static analyzers.

branches: 1.2.12; 1.2.30; 1.2.34; 1.2.38;
add dependency on "drm" module

branches: 1.1.2; 1.1.4; 1.1.8;
add VirtualBox DRM driver