ozone: drm: Allow Modeset to succeed even if cursor fails

ozone: drm: Allow Modeset to succeed even if cursor fails

This is a special case to allow QEMU to work. The QEMU
graphics driver is incomplete and does not provide a cursor.

BUG=471097
Committed: https://crrev.com/457659a06b2fe2e60005f48653492951d89cc409
Cr-Commit-Position: refs/heads/master@{#325654}

[spang, 2015-04-17 16:30:07]

spang@chromium.org changed reviewers:
+ dnicoara@chromium.org, dongseong.hwang@intel.com, zachr@chromium.org

[spang, 2015-04-17 16:30:07]

zachr: Can we have a cursor in qemu? :(

[dnicoara, 2015-04-17 16:31:28]

lgtm

[spang, 2015-04-17 16:31:59]

The CQ bit was checked by spang@chromium.org

[commit-bot: I haz the power, 2015-04-17 16:33:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1090263002/1

[zachr, 2015-04-17 16:38:12]

On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> CQ is trying da patch. Follow status at
>  https://chromium-cq-status.appspot.com/patch-status/1090263002/1

spang: I implemented it once before, but it had fundamental limitations due to
the emulated graphics hardware. It was decided that a software cursor would be
best.

[commit-bot: I haz the power, 2015-04-17 16:55:08]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-04-17 16:55:56]

Patchset 1 (id:??) landed as
https://crrev.com/457659a06b2fe2e60005f48653492951d89cc409
Cr-Commit-Position: refs/heads/master@{#325654}

[spang, 2015-04-17 17:02:40]

On 2015/04/17 16:38:12, zachr wrote:
> On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> > CQ is trying da patch. Follow status at
> >  https://chromium-cq-status.appspot.com/patch-status/1090263002/1
> 
> spang: I implemented it once before, but it had fundamental limitations due to
> the emulated graphics hardware. It was decided that a software cursor would be
> best.

Would the new virtio KMS driver help with that?

I think the virtualized hardware should not place extra requirements on
userspace. Otherwise it's not a good test (and will break, as demonstrated
here).

[zachr, 2015-04-17 18:01:56]

On 2015/04/17 17:02:40, spang wrote:
> On 2015/04/17 16:38:12, zachr wrote:
> > On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> > > CQ is trying da patch. Follow status at
> > >  https://chromium-cq-status.appspot.com/patch-status/1090263002/1
> > 
> > spang: I implemented it once before, but it had fundamental limitations due
to
> > the emulated graphics hardware. It was decided that a software cursor would
be
> > best.
> 
> Would the new virtio KMS driver help with that?
> 
> I think the virtualized hardware should not place extra requirements on
> userspace. Otherwise it's not a good test (and will break, as demonstrated
> here).

I agree. The virtio stuff would be helpful, but it's not done. That being said,
I haven't looked at the state of that project in a while. I'll take another look
at it.

[dshwang, 2015-04-20 15:14:22]

On 2015/04/17 16:38:12, zachr wrote:
> On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> > CQ is trying da patch. Follow status at
> >  https://chromium-cq-status.appspot.com/patch-status/1090263002/1
> 
> spang: I implemented it once before, but it had fundamental limitations due to
> the emulated graphics hardware. It was decided that a software cursor would be
> best.

zachr@, does it mean that you can see software cursor in qemu?
I cannot see any cursor in qemu, so I have hard time to test. Could you give me
tip how to see cursor?

[spang, 2015-04-20 15:20:08]

On 2015/04/20 15:14:22, dshwang wrote:
> On 2015/04/17 16:38:12, zachr wrote:
> > On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> > > CQ is trying da patch. Follow status at
> > >  https://chromium-cq-status.appspot.com/patch-status/1090263002/1
> > 
> > spang: I implemented it once before, but it had fundamental limitations due
to
> > the emulated graphics hardware. It was decided that a software cursor would
be
> > best.
> 
> zachr@, does it mean that you can see software cursor in qemu?
> I cannot see any cursor in qemu, so I have hard time to test. Could you give
me
> tip how to see cursor?

If you turn on the screen magnifier using the keyboard (under Accessibility)
you'll get a software cursor.

Without doing that, it's invisible.

We were going to enable SW cursor for DisplayLink (and qemu) but DisplayLink
doesn't need this any more. I don't think it's worthwhile to do all of the
necessary plumbing just for QEMU.

[zachr, 2015-04-20 17:07:44]

On 2015/04/20 15:20:08, spang wrote:
> On 2015/04/20 15:14:22, dshwang wrote:
> > On 2015/04/17 16:38:12, zachr wrote:
> > > On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> > > > CQ is trying da patch. Follow status at
> > > >  https://chromium-cq-status.appspot.com/patch-status/1090263002/1
> > > 
> > > spang: I implemented it once before, but it had fundamental limitations
due
> to
> > > the emulated graphics hardware. It was decided that a software cursor
would
> be
> > > best.
> > 
> > zachr@, does it mean that you can see software cursor in qemu?
> > I cannot see any cursor in qemu, so I have hard time to test. Could you give
> me
> > tip how to see cursor?
> 
> If you turn on the screen magnifier using the keyboard (under Accessibility)
> you'll get a software cursor.
> 
> Without doing that, it's invisible.
> 
> We were going to enable SW cursor for DisplayLink (and qemu) but DisplayLink
> doesn't need this any more. I don't think it's worthwhile to do all of the
> necessary plumbing just for QEMU.

You can also enable big cursor mode, or high contrast mode to see the SW cursor.

[dshwang, 2015-04-20 17:54:59]

On 2015/04/20 17:07:44, zachr wrote:
> On 2015/04/20 15:20:08, spang wrote:
> > On 2015/04/20 15:14:22, dshwang wrote:
> > > On 2015/04/17 16:38:12, zachr wrote:
> > > > On 2015/04/17 16:33:48, I haz the power (commit-bot) wrote:
> > > > > CQ is trying da patch. Follow status at
> > > > >  https://chromium-cq-status.appspot.com/patch-status/1090263002/1
> > > > 
> > > > spang: I implemented it once before, but it had fundamental limitations
> due
> > to
> > > > the emulated graphics hardware. It was decided that a software cursor
> would
> > be
> > > > best.
> > > 
> > > zachr@, does it mean that you can see software cursor in qemu?
> > > I cannot see any cursor in qemu, so I have hard time to test. Could you
give
> > me
> > > tip how to see cursor?
> > 
> > If you turn on the screen magnifier using the keyboard (under Accessibility)
> > you'll get a software cursor.
> > 
> > Without doing that, it's invisible.
> > 
> > We were going to enable SW cursor for DisplayLink (and qemu) but DisplayLink
> > doesn't need this any more. I don't think it's worthwhile to do all of the
> > necessary plumbing just for QEMU.
> 
> You can also enable big cursor mode, or high contrast mode to see the SW
cursor.

Thx. big cursor works well