This patch attempts to fix up shortcomings with the current calling
sequences.
1) There's a fastpath where no locking occurs and only io_mem_reserved is
called to obtain needed info for mapping. The fastpath is set per
memory type manager.
2) If the fastpath is disabled, io_mem_reserve and io_mem_free will be exactly
balanced and not called recursively for the same struct ttm_mem_reg.
3) Optionally the driver can choose to enable a per memory type manager LRU
eviction mechanism that, when io_mem_reserve returns -EAGAIN will attempt
to kill user-space mappings of memory in that manager to free up needed
resources
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Rather than having the driver supply the validation sequence, leave that
responsibility to TTM. This saves some confusion and a function argument.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Drastically reduce the number of spin lock / unlock operations by performing
unreserving and fencing under global locks.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jerome Glisse <j.glisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The bo lock used only to protect the bo sync object members, and since it
is a per bo lock, fencing a buffer list will see a lot of locks and unlocks.
Replace it with a per-device lock that protects the sync object members on
*all* bos. Reading and setting these members will always be very quick, so
the risc of heavy lock contention is microscopic. Note that waiting for
sync objects will always take place outside of this lock.
The bo device fence lock will eventually be replaced with a seqlock /
rcu mechanism so we can determine that a bo is idle under a
rcu / read seqlock.
However this change will allow us to batch fencing and unreserving of
buffers with a minimal amount of locking.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jerome Glisse <j.glisse@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Add an aid for the driver to detect deadlocks on multi-bo reservations
Update documentation.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jerome Glisse <j.glisse@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Makes it possible to reserve a list of buffer objects with a single
spin lock / unlock if there is no contention.
Should improve cpu usage on SMP kernels.
v2: Initialize private list members on reserve and don't call
ttm_bo_list_ref_sub() with zero put_count.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
A process suspended waiting for a higher sequence or no sequence to unreserve,
a bo may be beaten to the reservation by a process with a lower sequence.
In that case the first process should give up trying to reserve and
return -EAGAIN. In order for that to happen, we must wake waiting processes
when we change sequence, so that they have a chance to detect the new
sequence.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Call destroy() on _all_ ttm_bo_init() failures, and make sure that
behavior is documented in the function description.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This breaks vmwgfx non-root EGL clients and is a remnant from the
TTM user-space interface. This test should be done in the driver.
Replace the remaining placement test with a BUG_ON, since triggering
it is a driver bug.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The sync object may disappear as soon as we release the bo::lock, so
take a reference on it while we use it.
One option would be to call sync_object_flush() before releasing the bo::lock,
but that would put an atomic requirement on that function.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Since we're doing this outside of a spinlock to provide the necessary
barriers, add an explicit barrier.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Replace with BUG_ON(). These error messages remained from the time
when TTM was initialized from user-space. Nowadays hitting one of those
is really a kernel bug.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Remove an obsolete comment about mm nodes.
Document the new bo range manager interface.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This commit replaces the ttm_bo_cleanup_ref function with two new functions.
One for the case where the bo is not yet on the delayed destroy list, and
one for the case where the bo was on the delayed destroy list, at least at
the time of call. This makes it possible to optimize the two cases somewhat.
It also enables the possibility to directly destroy buffers on the
delayed delete list when they are about to be evicted or swapped out.
Currently they were only evicted / swapped and destruction was left for the
delayed buffer destruction thread.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Function ttm_bo_wait_unreserved can be slightly simplified.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
We need the unlocked variant for the new codepath introduced to fix the
race condition in master recently.
Signed-off-by: Dave Airlie <airlied@redhat.com>
This fixes a race pointed out by Dave Airlie where we don't take a buffer
object about to be destroyed off the LRU lists properly. It also fixes a rare
case where a buffer object could be destroyed in the middle of an
accelerated eviction.
The patch also adds a utility function that can be used to prematurely
release GPU memory space usage of an object waiting to be destroyed.
For example during eviction or swapout.
The above mentioned commit didn't queue the buffer on the delayed destroy
list under some rare circumstances. It also didn't completely honor the
remove_all parameter.
Fixes:
https://bugzilla.redhat.com/show_bug.cgi?id=615505http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=591061
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Nouveau will need this on GeForce 8 and up to account for the GPU
reordering physical VRAM for some memory types.
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Acked-by: Thomas Hellström <thellstrom@vmware.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Existing core code/drivers call drm_mm_put_block on ttm_mem_reg.mm_node
directly. Future patches will modify TTM behaviour in such a way that
ttm_mem_reg.mm_node doesn't necessarily belong to drm_mm.
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Acked-by: Thomas Hellström <thellstrom@vmware.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
I wrote this for the prime sharing work, but I also noticed other external
non-upstream drivers from a large company carrying a similiar patch, so I
may as well ship it in master.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Only ever assigned, never used.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[glisse: I will re-add if needed for range-restricted allocations]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Dave Airlie <airlied@redhat.com>
* 'drm-for-2.6.35' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6: (207 commits)
drm/radeon/kms/pm/r600: select the mid clock mode for single head low profile
drm/radeon: fix power supply kconfig interaction.
drm/radeon/kms: record object that have been list reserved
drm/radeon: AGP memory is only I/O if the aperture can be mapped by the CPU.
drm/radeon/kms: don't default display priority to high on rs4xx
drm/edid: fix typo in 1600x1200@75 mode
drm/nouveau: fix i2c-related init table handlers
drm/nouveau: support init table i2c device identifier 0x81
drm/nouveau: ensure we've parsed i2c table entry for INIT_*I2C* handlers
drm/nouveau: display error message for any failed init table opcode
drm/nouveau: fix init table handlers to return proper error codes
drm/nv50: support fractional feedback divider on newer chips
drm/nv50: fix monitor detection on certain chipsets
drm/nv50: store full dcb i2c entry from vbios
drm/nv50: fix suspend/resume with DP outputs
drm/nv50: output calculated crtc pll when debugging on
drm/nouveau: dump pll limits entries when debugging is on
drm/nouveau: bios parser fixes for eDP boards
drm/nouveau: fix a nouveau_bo dereference after it's been destroyed
drm/nv40: remove some completed ctxprog TODOs
...
We want to be able to prevent the delayed workqueue from changing state
while we're reclocking, so add an API to block and unblock it.
Signed-off-by: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
It's unused and buggy in its current form, since it can place a bo
in the reserved state without removing it from lru lists.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
All TTM driver have been converted to new io_mem_reserve/free
interface which allow driver to choose and return proper io
base, offset to core TTM for ioremapping if necessary. This
patch remove what is now deadcode.
V2 adapt to match with change in first patch of the patchset
V3 update after io_mem_reserve/io_mem_free callback balancing
V4 adjust to minor cleanup
V5 remove the needs ioremap flag
V6 keep the ioremapping facility in TTM
[airlied- squashed driver removals in here also]
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
On fault the driver is given the opportunity to perform any operation
it sees fit in order to place the buffer into a CPU visible area of
memory. This patch doesn't break TTM users, nouveau, vmwgfx and radeon
should keep working properly. Future patch will take advantage of this
infrastructure and remove the old path from TTM once driver are
converted.
V2 return VM_FAULT_NOPAGE if callback return -EBUSY or -ERESTARTSYS
V3 balance io_mem_reserve and io_mem_free call, fault_reserve_notify
is responsible to perform any necessary task for mapping to succeed
V4 minor cleanup, atomic_t -> bool as member is protected by reserve
mecanism from concurent access
V5 the callback is now responsible for iomapping the bo and providing
a virtual address this simplify TTM and will allow to get rid of
TTM_MEMTYPE_FLAG_NEEDS_IOREMAP
V6 use the bus addr data to decide to ioremap or this isn't needed
but we don't necesarily need to ioremap in the callback but still
allow driver to use static mapping
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
There is case where we want to be able to wait only for the
GPU while not waiting for other buffer to be unreserved. This
patch split the no_wait argument all the way down in the whole
ttm path so that upper level can decide on what to wait on or
not.
[airlied: squashed these 4 for bisectability reasons.]
drm/radeon/kms: update to TTM no_wait splitted argument
drm/nouveau: update to TTM no_wait splitted argument
drm/vmwgfx: update to TTM no_wait splitted argument
[vmwgfx patch: Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>]
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Acked-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Replace sequential calls to kobject_init() and kobject_add() with the
combo wrapper kobject_init_and_add(), which provides the same
semantics.
Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Constify struct sysfs_ops.
This is part of the ops structure constification
effort started by Arjan van de Ven et al.
Benefits of this constification:
* prevents modification of data that is shared
(referenced) by many other structure instances
at runtime
* detects/prevents accidental (but not intentional)
modification attempts on archs that enforce
read-only kernel data at runtime
* potentially better optimized code as the compiler
can assume that the const data cannot be changed
* the compiler/linker move const data into .rodata
and therefore exclude them from false sharing
Signed-off-by: Emese Revfy <re.emese@gmail.com>
Acked-by: David Teigland <teigland@redhat.com>
Acked-by: Matt Domsch <Matt_Domsch@dell.com>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Acked-by: Hans J. Koch <hjk@linutronix.de>
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Acked-by: Jens Axboe <jens.axboe@oracle.com>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
If the buffer object was already in the requested memory type, but
outside of the requested range it was never moved into the requested range.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is needed to fix a vmwgfx memory usage bug.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Resending this with Thomas Hellstrom's signoff for merging into 2.6.33
ttm_bo_delayed_delete has a race condition, because after we do:
kref_put(&nentry->list_kref, ttm_bo_release_list);
we are not holding the list lock and not holding any reference to
objects, and thus every bo in the list can be removed and freed at
this point.
However, we then use the next pointer we stored, which is not guaranteed
to be valid.
This was apparently the cause of some Nouveau oopses I experienced.
This patch rewrites the function so that it keeps the reference to nentry
until nentry itself is freed and we already got a reference to nentry->next.
v2 updated by me according to Thomas Hellstrom's feedback.
v3 proposed by Thomas Hellstrom. Commit comment updated by me.
Both updates fixed minor efficiency/style issues only and all three versions
should be correct.
Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is a convention that the vmwgfx driver has come to rely on.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is needed for a bugfix in the vmwgfx driver.
Drivers may have GPU bindings on buffers that core TTM is not aware of,
and TTM may view those buffers as ordinary system memory buffers.
Add a notifier to such drivers when TTM is about to move the buffer
contents out to swappable memory. The driver must then release any
private GPU bindings on those buffers.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
System memory type doesn't have a drm_mm manager associated to
it. This patch avoid trying to call drm_mm_debug on unitialized
drm_mm when printing debug info on the system memory manager.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
1) The function was previously called with a potentially empty
LRU list which would have lead to an OOPS or servere corruption.
2) In rare cases, after reservation has succeeded, another process may
already have evicted it or even pinned it. We must revalidate the
buffer status after releasing the lru lock.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
1) Remove from lru before reserving so we avoid competing with
evicting processes.
2) Avoid calling kref_put() on bo::list_kref while spinlocked.
3) Additional refcounting bug-checking.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
a) the loops were going to <= not <, leading to illegal memory access
b) the busy placement checks were using the placement arrays not the
busy placement ones.
Acked-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
These are functions required by nouveau which will be merged later.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Convert ttm_buffer_object_init to use struct ttm_placement and
rename to ttm_bo_init for consistency with function naming. This
allow to give more complex placement at buffer creation. For
instance you ask to allocate bo into vram first but if there is
not enough vram you can give system as a second possible
placement. It also allow to create buffer in a specific range.
Also rename ttm_buffer_object_validate to ttm_bo_validate.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This add helper function to print information on eviction placements
and memory manager status when eviction fails to allocate memory
space.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This would allow to catch driver callback error of not properly
setting the eviction placement structure.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
Return -ERESTARTSYS instead of -ERESTART when interrupted by a signal.
The -ERESTARTSYS is converted to an -EINTR by the kernel signal layer
before returned to user-space.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This change allow driver to pass sorted memory placement,
from most prefered placement to least prefered placement.
In order to avoid long function prototype a structure is
used to gather memory placement informations such as range
restriction (if you need a buffer to be in given range).
Range restriction is determined by fpfn & lpfn which are
the first page and last page number btw which allocation
can happen. If those fields are set to 0 ttm will assume
buffer can be put anywhere in the address space (thus it
avoids putting a burden on the driver to always properly
set those fields).
This patch also factor few functions like evicting first
entry of lru list or getting a memory space. This avoid
code duplication.
V2: Change API to use placement flags and array instead
of packing placement order into a quadword.
V3: Make sure we set the appropriate mem.placement flag
when validating or allocation memory space.
[Pending Thomas Hellstrom further review but okay
from preliminary review so far].
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
None of the in-tree drivers use user objects yet so this wasn't hitting
us.
Stanse found unreachable code in ttm_bo_add_ttm:
http://decibel.fi.muni.cz/~xslaby/stanse/error.cgi?db=32&id=714#l238
Reported-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
ttm:
Remove a stray debug printout.
Remove a re-init of the lru spinlock at device init.
radeon:
Fix the size of the bo_global allocation.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Common resources, like memory accounting and swap lists should be
global and not per device. Introduce a struct ttm_bo_global to
accomodate this, and register it with sysfs. Add a small sysfs interface
to return the number of active buffer objects.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
Use inclusive zones to simplify accounting and its sysfs representation.
Use DMA32 accounting where applicable.
Add a sysfs interface to make the heuristically determined limits
readable and configurable.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
Check whether index is within bounds before grabbing the element.
Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
If there are multiple simultaneous waiters for the same buffer object,
a temporary reference to its sync object may be leaked.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This adds new set/get tiling interfaces where the pitch
and macro/micro tiling enables can be set. Along with
a flag to decide if this object should have a surface when mapped.
The only thing we need to allocate with a mapped surface should be
the frontbuffer. Note rotate scanout shouldn't require one, and
back/depth shouldn't either, though mesa needs some fixes.
It fixes the TTM interfaces along Thomas's suggestions, and I've tested
the surface stealing code with two X servers and not seen any lockdep issues.
I've stopped tiling the fbcon frontbuffer, as I don't see there being
any advantage other than testing, I've left the testing commands in there,
just flip the fb_tiled to true in radeon_fb.c
Open: Can we integrate endian swapping in with this?
Future features:
texture tiling - need to relocate texture registers TXOFFSET* with tiling info.
This also merges Michel's cleanup surfaces regs at init time patch
even though it makes sense on its own, this patch really relies on it.
Some PowerMac firmwares set up a tiling surface at the beginning of VRAM
which messes us up otherwise.
that patch is:
Signed-off-by: Michel Dänzer <daenzer@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This add support for using dma32 memory on gpus that really need it.
Currently IGPs are left without DMA32 but we might need to change
that unless we can fix rs690.
Signed-off-by: Dave Airlie <airlied@redhat.com>
A bug caused a new caching state to be selected on each buffer object
validation regardless of the current caching state.
Moreover, a caching state could be selected that wasn't supported by
the memory type.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
A bug caused the ttm code to just terminate the wait when a signal
was received while waiting for the GPU to release a buffer object that
was to be evicted.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
TTM is a GPU memory manager subsystem designed for use with GPU
devices with various memory types (On-card VRAM, AGP,
PCI apertures etc.). It's essentially a helper library that assists
the DRM driver in creating and managing persistent buffer objects.
TTM manages placement of data and CPU map setup and teardown on
data movement. It can also optionally manage synchronization of
data on a per-buffer-object level.
TTM takes care to provide an always valid virtual user-space address
to a buffer object which makes user-space sub-allocation of
big buffer objects feasible.
TTM uses a fine-grained per buffer-object locking scheme, taking
care to release all relevant locks when waiting for the GPU.
Although this implies some locking overhead, it's probably a big
win for devices with multiple command submission mechanisms, since
the lock contention will be minimal.
TTM can be used with whatever user-space interface the driver
chooses, including GEM. It's used by the upcoming Radeon KMS DRM driver
and is also the GPU memory management core of various new experimental
DRM drivers.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Thomas Hellstrom