In Lucid, we were carrying the following 4 patches to disable KMS for a
known set of broken devices, specifically some i8xx chipsets [1]. We
purposely didn't carry these SAUCE patches forward to Maverick in hopes
that this would no longer be an issue. Unfortunately, this does not
appear to be the case. I'll be re-applying the following 4 patches to
Maverick.
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
08-31-2010, 05:27 PM
Tim Gardner
KMS Disablement for i8xx
On 08/31/2010 10:13 AM, leann.ogasawara@canonical.com wrote:
> In Lucid, we were carrying the following 4 patches to disable KMS for a
> known set of broken devices, specifically some i8xx chipsets [1]. We
> purposely didn't carry these SAUCE patches forward to Maverick in hopes
> that this would no longer be an issue. Unfortunately, this does not
> appear to be the case. I'll be re-applying the following 4 patches to
> Maverick.
>
> Thanks,
> Leann
>
> [1] https://wiki.ubuntu.com/X/Bugs/Lucidi8xxFreezes
>
> Andy Whitcroft (4):
> UBUNTU: SAUCE: i915 KMS -- support disabling KMS for known broken
> devices
> UBUNTU: SAUCE: i915 KMS -- blacklist i830
> UBUNTU: SAUCE: i915 KMS -- blacklist i845g
> UBUNTU: SAUCE: i915 KMS -- blacklist i855
>
> drivers/gpu/drm/i915/i915_drv.c | 16 ++++++++++++++++
> 1 files changed, 16 insertions(+), 0 deletions(-)
>
>
Acked-by: Tim Gardner <tim.gardner@canonical.com>
--
Tim Gardner tim.gardner@canonical.com
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-02-2010, 06:25 PM
Steven
KMS Disablement for i8xx
Dear Kernel Team,
Please do NOT reapply these patches. UMS causes the majority of 8xx cards to boot to a black screen on Lucid and forward. I spent at least a month fixing probably a hundred different peoples' install problems after Lucid was released because of these patches.
If you do decide to reapply these, PLEASE consider paying especially close attention to bugs related to this change, at the very least.
I realize support on i8xx is difficult, but please be aware that these patches caused more harm than good for the Lucid release. Thank you for your time.
-Stenten
On Tue, Aug 31, 2010 at 1:27 PM, Tim Gardner <tim.gardner@canonical.com> wrote:
On 08/31/2010 10:13 AM, leann.ogasawara@canonical.com wrote:
> In Lucid, we were carrying the following 4 patches to disable KMS for a
> known set of broken devices, specifically some i8xx chipsets [1]. *We
> purposely didn't carry these SAUCE patches forward to Maverick in hopes
> that this would no longer be an issue. *Unfortunately, this does not
> appear to be the case. *I'll be re-applying the following 4 patches to
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-02-2010, 07:18 PM
Leann Ogasawara
KMS Disablement for i8xx
On Thu, 2010-09-02 at 14:25 -0400, Steven wrote:
> Dear Kernel Team,
>
>
> Please do NOT reapply these patches. UMS causes the majority of 8xx
> cards to boot to a black screen on Lucid and forward. I spent at least
> a month fixing probably a hundred different peoples' install problems
> after Lucid was released because of these patches.
>
>
> If you do decide to reapply these, PLEASE consider paying especially
> close attention to bugs related to this change, at the very least.
>
>
> I realize support on i8xx is difficult, but please be aware that these
> patches caused more harm than good for the Lucid release. Thank you
> for your time.
Thanks for the email Stenten.
Chris, you were the origination for the request to re-apply these KMS
disablement patches to Maverick. What's your take on Stenten's comments
above. I haven't yet re-applied these to Maverick and will hold off
until we hear back from you.
Thanks,
Leann
> On Tue, Aug 31, 2010 at 1:27 PM, Tim Gardner
> <tim.gardner@canonical.com> wrote:
> On 08/31/2010 10:13 AM, leann.ogasawara@canonical.com wrote:
> > In Lucid, we were carrying the following 4 patches to
> disable KMS for a
> > known set of broken devices, specifically some i8xx chipsets
> [1]. We
> > purposely didn't carry these SAUCE patches forward to
> Maverick in hopes
> > that this would no longer be an issue. Unfortunately, this
> does not
> > appear to be the case. I'll be re-applying the following 4
> patches to
> > Maverick.
> >
> > Thanks,
> > Leann
> >
> > [1] https://wiki.ubuntu.com/X/Bugs/Lucidi8xxFreezes
> >
> > Andy Whitcroft (4):
> > UBUNTU: SAUCE: i915 KMS -- support disabling KMS for
> known broken
> > devices
> > UBUNTU: SAUCE: i915 KMS -- blacklist i830
> > UBUNTU: SAUCE: i915 KMS -- blacklist i845g
> > UBUNTU: SAUCE: i915 KMS -- blacklist i855
> >
> > drivers/gpu/drm/i915/i915_drv.c | 16 ++++++++++++++++
> > 1 files changed, 16 insertions(+), 0 deletions(-)
> >
> >
>
>
> Acked-by: Tim Gardner <tim.gardner@canonical.com>
>
> --
> Tim Gardner tim.gardner@canonical.com
>
>
> --
> kernel-team mailing list
> kernel-team@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team
>
>
>
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-02-2010, 11:52 PM
Christopher James Halse Rogers
KMS Disablement for i8xx
On Thu, 2010-09-02 at 14:25 -0400, Steven wrote:
> Dear Kernel Team,
>
>
> Please do NOT reapply these patches. UMS causes the majority of 8xx
> cards to boot to a black screen on Lucid and forward. I spent at least
> a month fixing probably a hundred different peoples' install problems
> after Lucid was released because of these patches.
>
The situation in Maverick is different. In Lucid, we had the 2.9.1
Intel X driver, which still had a UMS codepath. In Maverick, we have
2.12, which does not have any UMS code. The Intel driver in Maverick
will not load without KMS.
Disabling KMS on these cards in Lucid resulted in them using the UMS
codepaths of the Intel X driver, and in many cases failing.
Disabling KMS on these cards in Maverick will result in them using the
VESA X driver, which has been reported to work on the various i8xx bugs.
>
> If you do decide to reapply these, PLEASE consider paying especially
> close attention to bugs related to this change, at the very least.
>
I will.
>
> I realize support on i8xx is difficult, but please be aware that these
> patches caused more harm than good for the Lucid release. Thank you
> for your time.
>
>
> -Stenten
>
>
> On Tue, Aug 31, 2010 at 1:27 PM, Tim Gardner
> <tim.gardner@canonical.com> wrote:
> On 08/31/2010 10:13 AM, leann.ogasawara@canonical.com wrote:
> > In Lucid, we were carrying the following 4 patches to
> disable KMS for a
> > known set of broken devices, specifically some i8xx chipsets
> [1]. We
> > purposely didn't carry these SAUCE patches forward to
> Maverick in hopes
> > that this would no longer be an issue. Unfortunately, this
> does not
> > appear to be the case. I'll be re-applying the following 4
> patches to
> > Maverick.
> >
> > Thanks,
> > Leann
> >
> > [1] https://wiki.ubuntu.com/X/Bugs/Lucidi8xxFreezes
> >
> > Andy Whitcroft (4):
> > UBUNTU: SAUCE: i915 KMS -- support disabling KMS for
> known broken
> > devices
> > UBUNTU: SAUCE: i915 KMS -- blacklist i830
> > UBUNTU: SAUCE: i915 KMS -- blacklist i845g
> > UBUNTU: SAUCE: i915 KMS -- blacklist i855
> >
> > drivers/gpu/drm/i915/i915_drv.c | 16 ++++++++++++++++
> > 1 files changed, 16 insertions(+), 0 deletions(-)
> >
> >
>
>
> Acked-by: Tim Gardner <tim.gardner@canonical.com>
>
> --
> Tim Gardner tim.gardner@canonical.com
>
>
> --
> kernel-team mailing list
> kernel-team@lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team
>
>
>
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-03-2010, 12:10 AM
Leann Ogasawara
KMS Disablement for i8xx
On Fri, 2010-09-03 at 09:52 +1000, Christopher James Halse Rogers wrote:
> On Thu, 2010-09-02 at 14:25 -0400, Steven wrote:
> > Dear Kernel Team,
> >
> >
> > Please do NOT reapply these patches. UMS causes the majority of 8xx
> > cards to boot to a black screen on Lucid and forward. I spent at least
> > a month fixing probably a hundred different peoples' install problems
> > after Lucid was released because of these patches.
> >
>
> The situation in Maverick is different. In Lucid, we had the 2.9.1
> Intel X driver, which still had a UMS codepath. In Maverick, we have
> 2.12, which does not have any UMS code. The Intel driver in Maverick
> will not load without KMS.
>
> Disabling KMS on these cards in Lucid resulted in them using the UMS
> codepaths of the Intel X driver, and in many cases failing.
>
> Disabling KMS on these cards in Maverick will result in them using the
> VESA X driver, which has been reported to work on the various i8xx bugs.
>
> >
> > If you do decide to reapply these, PLEASE consider paying especially
> > close attention to bugs related to this change, at the very least.
> >
>
> I will.
Chris, thanks for the follow up and the clarification. Based on your
feedback, I'll go ahead and re-apply these patches.
Thanks,
Leann
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-03-2010, 01:42 AM
Leann Ogasawara
KMS Disablement for i8xx
Applied to Maverick linux master.
Thanks,
Leann
On Tue, 2010-08-31 at 09:13 -0700, leann.ogasawara@canonical.com wrote:
> In Lucid, we were carrying the following 4 patches to disable KMS for a
> known set of broken devices, specifically some i8xx chipsets [1]. We
> purposely didn't carry these SAUCE patches forward to Maverick in hopes
> that this would no longer be an issue. Unfortunately, this does not
> appear to be the case. I'll be re-applying the following 4 patches to
> Maverick.
>
> Thanks,
> Leann
>
> [1] https://wiki.ubuntu.com/X/Bugs/Lucidi8xxFreezes
>
> Andy Whitcroft (4):
> UBUNTU: SAUCE: i915 KMS -- support disabling KMS for known broken
> devices
> UBUNTU: SAUCE: i915 KMS -- blacklist i830
> UBUNTU: SAUCE: i915 KMS -- blacklist i845g
> UBUNTU: SAUCE: i915 KMS -- blacklist i855
>
> drivers/gpu/drm/i915/i915_drv.c | 16 ++++++++++++++++
> 1 files changed, 16 insertions(+), 0 deletions(-)
>
>
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-05-2010, 06:09 PM
Matthew Garrett
KMS Disablement for i8xx
On Thu, Sep 02, 2010 at 05:10:11PM -0700, Leann Ogasawara wrote:
> On Fri, 2010-09-03 at 09:52 +1000, Christopher James Halse Rogers wrote:
> > Disabling KMS on these cards in Maverick will result in them using the
> > VESA X driver, which has been reported to work on the various i8xx bugs.
It's probably worth noting that a large number of 855 machines have
native panel resolutions that aren't present in the vesa tables, and so
won't be able to run in native resolution with the vesa driver.
--
Matthew Garrett | mjg59@srcf.ucam.org
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-05-2010, 09:05 PM
Steven
KMS Disablement for i8xx
On Sun, Sep 5, 2010 at 2:09 PM, Matthew Garrett <mjg59@srcf.ucam.org> wrote:
On Thu, Sep 02, 2010 at 05:10:11PM -0700, Leann Ogasawara wrote:
> On Fri, 2010-09-03 at 09:52 +1000, Christopher James Halse Rogers wrote:
> > Disabling KMS on these cards in Maverick will result in them using the
> > VESA X driver, which has been reported to work on the various i8xx bugs.
It's probably worth noting that a large number of 855 machines have
native panel resolutions that aren't present in the vesa tables, and so
won't be able to run in native resolution with the vesa driver.
This has been my experience. When running the vesa driver on Lucid on my i855, I getEE) VESA: No valid modes.
(EE) Screen(s) found, but none have a usable configuration.Then the graphics never load.
-Stenten
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
09-08-2010, 08:46 AM
Brian Rogers
KMS Disablement for i8xx
On 09/05/2010 02:05 PM, Steven wrote:
On Sun, Sep 5, 2010 at 2:09 PM, Matthew Garrett <mjg59@srcf.ucam.org>
wrote:
On Thu, Sep 02, 2010 at 05:10:11PM -0700,
Leann Ogasawara wrote:
> On Fri, 2010-09-03 at 09:52 +1000, Christopher James
Halse Rogers wrote:
> > Disabling KMS on these cards in
Maverick will result in them using the
> > VESA X driver, which has been reported to work on
the various i8xx bugs.
It's probably worth noting that a large number of 855 machines
have
native panel resolutions that aren't present in the vesa
tables, and so
won't be able to run in native resolution with the vesa
driver.
This has been my experience. When running the vesa driver
on Lucid on my i855, I get:
(EE) VESA: No valid modes.
(EE) Screen(s) found, but none have a usable configuration.
Then the graphics never load.
Wouldn't the fbdev X driver be a better option? I don't believe it
suffers from the same stability issues, and it should always be able
to use the native panel resolution. I'd also trust it more with
suspend/resume and VT or user switching. Does vesa have any
advantages over fbdev?
To support fbdev fallback for problem cards, the
xserver-xorg-video-intel driver would have to be patched to
blacklist i8xx cards (with support for overriding in xorg.conf), and
we'd have to make sure fbdev is tried as a fallback before vesa.
--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
KMS Disablement for i8xx
EE) VESA: No valid modes.
