<feed xmlns='http://www.w3.org/2005/Atom'>
<title>kernel.git/drivers/gpu/drm/ttm/ttm_execbuf_util.c, branch linux-3.14.y</title>
<subtitle>Hosts the 0x221E linux distro kernel.
</subtitle>
<id>https://git.0xinfinity.dev/distro/kernel.git/atom?h=linux-3.14.y</id>
<link rel='self' href='https://git.0xinfinity.dev/distro/kernel.git/atom?h=linux-3.14.y'/>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/'/>
<updated>2013-11-18T08:38:52Z</updated>
<entry>
<title>drm/ttm: Allow execbuf util reserves without ticket</title>
<updated>2013-11-18T08:38:52Z</updated>
<author>
<name>Thomas Hellstrom</name>
</author>
<published>2013-11-15T08:02:54Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=8d17fb4455eadbf2642e34e5584d0b0de64d33a8'/>
<id>urn:sha1:8d17fb4455eadbf2642e34e5584d0b0de64d33a8</id>
<content type='text'>
If no reservation ticket is given to the execbuf reservation utilities,
try reservation with non-blocking semantics.
This is intended for eviction paths that use the execbuf reservation
utilities for convenience rather than for deadlock avoidance.

Signed-off-by: Thomas Hellstrom &lt;thellstrom@vmware.com&gt;
Reviewed-by: Jakob Bornecrantz &lt;jakob@vmware.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: inline ttm_bo_reserve and related calls</title>
<updated>2013-06-28T02:04:09Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2013-06-27T11:48:24Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=3482032457f50cae196f6397ebec7f5f2ad3cf7d'/>
<id>urn:sha1:3482032457f50cae196f6397ebec7f5f2ad3cf7d</id>
<content type='text'>
Makes lockdep a lot more useful.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Signed-off-by: Dave Airlie &lt;airlied@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: convert to the reservation api</title>
<updated>2013-06-28T02:04:01Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2013-06-27T11:48:19Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=5e338405119a80aa59e811626739122d1c15045d'/>
<id>urn:sha1:5e338405119a80aa59e811626739122d1c15045d</id>
<content type='text'>
Now that the code is compatible in semantics, flip the switch.
Use ww_mutex instead of the homegrown implementation.

ww_mutex uses -EDEADLK to signal that the caller has to back off,
and -EALREADY to indicate this buffer is already held by the caller.

ttm used -EAGAIN and -EDEADLK for those, respectively. So some changes
were needed to handle this correctly.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Jerome Glisse &lt;jglisse@redhat.com&gt;
Signed-off-by: Dave Airlie &lt;airlied@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: make ttm reservation calls behave like reservation calls</title>
<updated>2013-06-28T02:02:20Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2013-06-27T11:48:17Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=ecff665f5e3f1c6909353e00b9420e45ae23d995'/>
<id>urn:sha1:ecff665f5e3f1c6909353e00b9420e45ae23d995</id>
<content type='text'>
This commit converts the source of the val_seq counter to
the ww_mutex api. The reservation objects are converted later,
because there is still a lockdep splat in nouveau that has to
resolved first.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Jerome Glisse &lt;jglisse@redhat.com&gt;
Signed-off-by: Dave Airlie &lt;airlied@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: use ttm_bo_reserve_slowpath_nolru in ttm_eu_reserve_buffers, v2</title>
<updated>2013-01-15T13:57:10Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2013-01-15T13:57:10Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=f2d476a110bc24fde008698ae9018c99e803e25c'/>
<id>urn:sha1:f2d476a110bc24fde008698ae9018c99e803e25c</id>
<content type='text'>
This requires re-use of the seqno, which increases fairness slightly.
Instead of spinning with a new seqno every time we keep the current one,
but still drop all other reservations we hold. Only when we succeed,
we try to get back our other reservations again.

This should increase fairness slightly as well.

Changes since v1:
 - Increase val_seq before calling ttm_bo_reserve_slowpath_nolru and
   retrying to take all entries to prevent a race.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Jerome Glisse &lt;jglisse@redhat.com&gt;

</content>
</entry>
<entry>
<title>drm/ttm: cleanup ttm_eu_reserve_buffers handling</title>
<updated>2013-01-15T13:56:48Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2013-01-15T13:56:48Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=7a1863084c9d90ce4b67d645bf9b0f1612e68f62'/>
<id>urn:sha1:7a1863084c9d90ce4b67d645bf9b0f1612e68f62</id>
<content type='text'>
With the lru lock no longer required for protecting reservations we
can just do a ttm_bo_reserve_nolru on -EBUSY, and handle all errors
in a single path.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Jerome Glisse &lt;jglisse@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: remove lru_lock around ttm_bo_reserve</title>
<updated>2013-01-15T13:56:37Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2013-01-15T13:56:37Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=63d0a4195560362e2e00a3ad38fc331d34e1da9b'/>
<id>urn:sha1:63d0a4195560362e2e00a3ad38fc331d34e1da9b</id>
<content type='text'>
There should no longer be assumptions that reserve will always succeed
with the lru lock held, so we can safely break the whole atomic
reserve/lru thing. As a bonus this fixes most lockdep annotations for
reservations.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Jerome Glisse &lt;jglisse@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: change fence_lock to inner lock</title>
<updated>2012-12-10T10:09:58Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2012-11-28T11:25:39Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=4154f051e74e6a5db174c8f4fc8a2f9c8a6b2541'/>
<id>urn:sha1:4154f051e74e6a5db174c8f4fc8a2f9c8a6b2541</id>
<content type='text'>
This requires changing the order in ttm_bo_cleanup_refs_or_queue to
take the reservation first, as there is otherwise no race free way to
take lru lock before fence_lock.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Thomas Hellstrom &lt;thellstrom@vmware.com&gt;
Signed-off-by: Dave Airlie &lt;airlied@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: alter cpu_writers to return -EBUSY in ttm_execbuf_util reservations</title>
<updated>2012-11-20T06:17:35Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2012-11-06T13:39:43Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=654aa79259a19f0d5e3cf9cb20aff56dc3b041b7'/>
<id>urn:sha1:654aa79259a19f0d5e3cf9cb20aff56dc3b041b7</id>
<content type='text'>
This is similar to other platforms that don't allow command submission
to buffers locked on the cpu.

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-by: Thomas Hellstrom &lt;thellstrom@vmware.com&gt;
Signed-off-by: Dave Airlie &lt;airlied@redhat.com&gt;
</content>
</entry>
<entry>
<title>drm/ttm: remove sync_obj_arg member</title>
<updated>2012-11-20T06:09:55Z</updated>
<author>
<name>Maarten Lankhorst</name>
</author>
<published>2012-10-12T15:02:19Z</published>
<link rel='alternate' type='text/html' href='https://git.0xinfinity.dev/distro/kernel.git/commit/?id=5fb4ef0e36b4c6ecc7fb025aaacb3b63b1114e87'/>
<id>urn:sha1:5fb4ef0e36b4c6ecc7fb025aaacb3b63b1114e87</id>
<content type='text'>
vmwgfx was its only user and always sets it to the same..

Signed-off-by: Maarten Lankhorst &lt;maarten.lankhorst@canonical.com&gt;
Reviewed-By: Thomas Hellstrom &lt;thellstrom@vmware.com&gt;
Signed-off-by: Dave Airlie &lt;airlied@redhat.com&gt;
</content>
</entry>
</feed>
