drm-next + i915 CVE yolo merge

drm-next + i915 CVE yolo merge

[Dave Airlie]

The landing of the i915 CVE fixes into Linus tree has created a bit of
a mess in linux-next and downstream in drm-next trees.

I talked to Daniel and he had talked to Joonas a bit, and I decided to
go with what Daniel describes as the YOLO merge, where I just solve it
and pray, and everyone else verifies/fixes it.

In my favour I've been reading these patches for a couple of months
now and applied them to a lot of places, so I'm quite familiar with
what they are doing.

The worst culprit was the RC6 ctx corruption fix since the whole of
rc6 got refactored in next. However I also had access to a version of
this patch Jon wrote on drm-tip a couple of weeks ago.

I took that patch, applied it and fixed it up on top of drm-next. Then
I backmerged the commit that also went into Linus' tree. Then I
removed any evidence of the RC6 patch from Linus' tree and left the
newer version pieces in place. The other stuff mostly merged fine and
the results looked fine, but I'd definitely think everyone at Intel
should be staring at it, and my dinq tip resolutions ASAP and
hopefully it goes into CI and comes out smelling of something good.

Let me know if it's all horrible asap,
Thanks,
Dave.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

drm-next + i915 CVE yolo merge

[Dave Airlie]

The landing of the i915 CVE fixes into Linus tree has created a bit of
a mess in linux-next and downstream in drm-next trees.

I talked to Daniel and he had talked to Joonas a bit, and I decided to
go with what Daniel describes as the YOLO merge, where I just solve it
and pray, and everyone else verifies/fixes it.

In my favour I've been reading these patches for a couple of months
now and applied them to a lot of places, so I'm quite familiar with
what they are doing.

The worst culprit was the RC6 ctx corruption fix since the whole of
rc6 got refactored in next. However I also had access to a version of
this patch Jon wrote on drm-tip a couple of weeks ago.

I took that patch, applied it and fixed it up on top of drm-next. Then
I backmerged the commit that also went into Linus' tree. Then I
removed any evidence of the RC6 patch from Linus' tree and left the
newer version pieces in place. The other stuff mostly merged fine and
the results looked fine, but I'd definitely think everyone at Intel
should be staring at it, and my dinq tip resolutions ASAP and
hopefully it goes into CI and comes out smelling of something good.

Let me know if it's all horrible asap,
Thanks,
Dave.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: drm-next + i915 CVE yolo merge

[Daniel Vetter]

On Thu, Nov 14, 2019 at 2:33 AM Dave Airlie <airlied@gmail.com> wrote:
> The landing of the i915 CVE fixes into Linus tree has created a bit of
> a mess in linux-next and downstream in drm-next trees.
>
> I talked to Daniel and he had talked to Joonas a bit, and I decided to
> go with what Daniel describes as the YOLO merge, where I just solve it
> and pray, and everyone else verifies/fixes it.
>
> In my favour I've been reading these patches for a couple of months
> now and applied them to a lot of places, so I'm quite familiar with
> what they are doing.
>
> The worst culprit was the RC6 ctx corruption fix since the whole of
> rc6 got refactored in next. However I also had access to a version of
> this patch Jon wrote on drm-tip a couple of weeks ago.
>
> I took that patch, applied it and fixed it up on top of drm-next. Then
> I backmerged the commit that also went into Linus' tree. Then I
> removed any evidence of the RC6 patch from Linus' tree and left the
> newer version pieces in place. The other stuff mostly merged fine and
> the results looked fine, but I'd definitely think everyone at Intel
> should be staring at it, and my dinq tip resolutions ASAP and
> hopefully it goes into CI and comes out smelling of something good.
>
> Let me know if it's all horrible asap,

Add Martin and Arek for CI results. The yolo stuff landed in
CI_DRM_7340, did we break anything in there? From a quick look seems
all ok.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: drm-next + i915 CVE yolo merge

[Bloomfield, Jon]

> -----Original Message-----
> From: Daniel Vetter <daniel.vetter@ffwll.ch>
> Sent: Thursday, November 14, 2019 12:13 AM
> To: Dave Airlie <airlied@gmail.com>
> Cc: Jani Nikula <jani.nikula@linux.intel.com>; Bloomfield, Jon
> <jon.bloomfield@intel.com>; Joonas Lahtinen
> <joonas.lahtinen@linux.intel.com>; Chris Wilson <chris@chris-wilson.co.uk>;
> Stephen Rothwell <sfr@canb.auug.org.au>; dri-devel <dri-
> devel@lists.freedesktop.org>; Intel Graphics Development <intel-
> gfx@lists.freedesktop.org>; Linus Torvalds <torvalds@linux-foundation.org>
> Subject: Re: drm-next + i915 CVE yolo merge
> 
> On Thu, Nov 14, 2019 at 2:33 AM Dave Airlie <airlied@gmail.com> wrote:
> > The landing of the i915 CVE fixes into Linus tree has created a bit of
> > a mess in linux-next and downstream in drm-next trees.
> >
> > I talked to Daniel and he had talked to Joonas a bit, and I decided to
> > go with what Daniel describes as the YOLO merge, where I just solve it
> > and pray, and everyone else verifies/fixes it.
> >
> > In my favour I've been reading these patches for a couple of months
> > now and applied them to a lot of places, so I'm quite familiar with
> > what they are doing.
> >
> > The worst culprit was the RC6 ctx corruption fix since the whole of
> > rc6 got refactored in next. However I also had access to a version of
> > this patch Jon wrote on drm-tip a couple of weeks ago.
> >
> > I took that patch, applied it and fixed it up on top of drm-next. Then
> > I backmerged the commit that also went into Linus' tree. Then I
> > removed any evidence of the RC6 patch from Linus' tree and left the
> > newer version pieces in place. The other stuff mostly merged fine and
> > the results looked fine, but I'd definitely think everyone at Intel
> > should be staring at it, and my dinq tip resolutions ASAP and
> > hopefully it goes into CI and comes out smelling of something good.
> >

Imre should look at the RC6 fix - He did all the hard work on that, including the rebases I sent to Dave. I was just a proxy :-)

> > Let me know if it's all horrible asap,
> 
> Add Martin and Arek for CI results. The yolo stuff landed in
> CI_DRM_7340, did we break anything in there? From a quick look seems
> all ok.

Jon
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

RE: drm-next + i915 CVE yolo merge

[Bloomfield, Jon]

> -----Original Message-----
> From: Daniel Vetter <daniel.vetter@ffwll.ch>
> Sent: Thursday, November 14, 2019 12:13 AM
> To: Dave Airlie <airlied@gmail.com>
> Cc: Jani Nikula <jani.nikula@linux.intel.com>; Bloomfield, Jon
> <jon.bloomfield@intel.com>; Joonas Lahtinen
> <joonas.lahtinen@linux.intel.com>; Chris Wilson <chris@chris-wilson.co.uk>;
> Stephen Rothwell <sfr@canb.auug.org.au>; dri-devel <dri-
> devel@lists.freedesktop.org>; Intel Graphics Development <intel-
> gfx@lists.freedesktop.org>; Linus Torvalds <torvalds@linux-foundation.org>
> Subject: Re: drm-next + i915 CVE yolo merge
> 
> On Thu, Nov 14, 2019 at 2:33 AM Dave Airlie <airlied@gmail.com> wrote:
> > The landing of the i915 CVE fixes into Linus tree has created a bit of
> > a mess in linux-next and downstream in drm-next trees.
> >
> > I talked to Daniel and he had talked to Joonas a bit, and I decided to
> > go with what Daniel describes as the YOLO merge, where I just solve it
> > and pray, and everyone else verifies/fixes it.
> >
> > In my favour I've been reading these patches for a couple of months
> > now and applied them to a lot of places, so I'm quite familiar with
> > what they are doing.
> >
> > The worst culprit was the RC6 ctx corruption fix since the whole of
> > rc6 got refactored in next. However I also had access to a version of
> > this patch Jon wrote on drm-tip a couple of weeks ago.
> >
> > I took that patch, applied it and fixed it up on top of drm-next. Then
> > I backmerged the commit that also went into Linus' tree. Then I
> > removed any evidence of the RC6 patch from Linus' tree and left the
> > newer version pieces in place. The other stuff mostly merged fine and
> > the results looked fine, but I'd definitely think everyone at Intel
> > should be staring at it, and my dinq tip resolutions ASAP and
> > hopefully it goes into CI and comes out smelling of something good.
> >

Imre should look at the RC6 fix - He did all the hard work on that, including the rebases I sent to Dave. I was just a proxy :-)

> > Let me know if it's all horrible asap,
> 
> Add Martin and Arek for CI results. The yolo stuff landed in
> CI_DRM_7340, did we break anything in there? From a quick look seems
> all ok.

Jon
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: drm-next + i915 CVE yolo merge

[Joonas Lahtinen]

Quoting Dave Airlie (2019-11-14 03:33:24)
> The landing of the i915 CVE fixes into Linus tree has created a bit of
> a mess in linux-next and downstream in drm-next trees.
> 
> I talked to Daniel and he had talked to Joonas a bit, and I decided to
> go with what Daniel describes as the YOLO merge, where I just solve it
> and pray, and everyone else verifies/fixes it.
> 
> In my favour I've been reading these patches for a couple of months
> now and applied them to a lot of places, so I'm quite familiar with
> what they are doing.
> 
> The worst culprit was the RC6 ctx corruption fix since the whole of
> rc6 got refactored in next. However I also had access to a version of
> this patch Jon wrote on drm-tip a couple of weeks ago.

We've now tested drm-next and found it working fine. As a next step Jani
will be backmerge to drm-intel-next-queued.

Regards, Joonas

> I took that patch, applied it and fixed it up on top of drm-next. Then
> I backmerged the commit that also went into Linus' tree. Then I
> removed any evidence of the RC6 patch from Linus' tree and left the
> newer version pieces in place. The other stuff mostly merged fine and
> the results looked fine, but I'd definitely think everyone at Intel
> should be staring at it, and my dinq tip resolutions ASAP and
> hopefully it goes into CI and comes out smelling of something good.
> 
> Let me know if it's all horrible asap,
> Thanks,
> Dave.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: drm-next + i915 CVE yolo merge

[Jani Nikula]

On Fri, 15 Nov 2019, Joonas Lahtinen <joonas.lahtinen@linux.intel.com> wrote:
> Quoting Dave Airlie (2019-11-14 03:33:24)
>> The landing of the i915 CVE fixes into Linus tree has created a bit of
>> a mess in linux-next and downstream in drm-next trees.
>> 
>> I talked to Daniel and he had talked to Joonas a bit, and I decided to
>> go with what Daniel describes as the YOLO merge, where I just solve it
>> and pray, and everyone else verifies/fixes it.
>> 
>> In my favour I've been reading these patches for a couple of months
>> now and applied them to a lot of places, so I'm quite familiar with
>> what they are doing.
>> 
>> The worst culprit was the RC6 ctx corruption fix since the whole of
>> rc6 got refactored in next. However I also had access to a version of
>> this patch Jon wrote on drm-tip a couple of weeks ago.
>
> We've now tested drm-next and found it working fine. As a next step Jani
> will be backmerge to drm-intel-next-queued.

Done and pushed out.

Thanks,
Jani.


>
> Regards, Joonas
>
>> I took that patch, applied it and fixed it up on top of drm-next. Then
>> I backmerged the commit that also went into Linus' tree. Then I
>> removed any evidence of the RC6 patch from Linus' tree and left the
>> newer version pieces in place. The other stuff mostly merged fine and
>> the results looked fine, but I'd definitely think everyone at Intel
>> should be staring at it, and my dinq tip resolutions ASAP and
>> hopefully it goes into CI and comes out smelling of something good.
>> 
>> Let me know if it's all horrible asap,
>> Thanks,
>> Dave.

-- 
Jani Nikula, Intel Open Source Graphics Center
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: drm-next + i915 CVE yolo merge

[Jani Nikula]

On Fri, 15 Nov 2019, Joonas Lahtinen <joonas.lahtinen@linux.intel.com> wrote:
> Quoting Dave Airlie (2019-11-14 03:33:24)
>> The landing of the i915 CVE fixes into Linus tree has created a bit of
>> a mess in linux-next and downstream in drm-next trees.
>> 
>> I talked to Daniel and he had talked to Joonas a bit, and I decided to
>> go with what Daniel describes as the YOLO merge, where I just solve it
>> and pray, and everyone else verifies/fixes it.
>> 
>> In my favour I've been reading these patches for a couple of months
>> now and applied them to a lot of places, so I'm quite familiar with
>> what they are doing.
>> 
>> The worst culprit was the RC6 ctx corruption fix since the whole of
>> rc6 got refactored in next. However I also had access to a version of
>> this patch Jon wrote on drm-tip a couple of weeks ago.
>
> We've now tested drm-next and found it working fine. As a next step Jani
> will be backmerge to drm-intel-next-queued.

Done and pushed out.

Thanks,
Jani.


>
> Regards, Joonas
>
>> I took that patch, applied it and fixed it up on top of drm-next. Then
>> I backmerged the commit that also went into Linus' tree. Then I
>> removed any evidence of the RC6 patch from Linus' tree and left the
>> newer version pieces in place. The other stuff mostly merged fine and
>> the results looked fine, but I'd definitely think everyone at Intel
>> should be staring at it, and my dinq tip resolutions ASAP and
>> hopefully it goes into CI and comes out smelling of something good.
>> 
>> Let me know if it's all horrible asap,
>> Thanks,
>> Dave.

-- 
Jani Nikula, Intel Open Source Graphics Center
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel