2009-12-20 16:23:16

by Johannes Hirte

[permalink] [raw]
Subject: [stable] Regression in 2.6.32.2: radeon KMS hangs system

With 2.6.32.2 the system hangs completely with black screen when modprobing
radeon with modeset=1. I was able to connect via ssh, but after entering the
password nothing happend. I didn't got a prompt. Only SysRq worked. I've
tracked it down to the following patch:

>From 500b758725314ab1b5316eb0caa5b0fa26740e6b Mon Sep 17 00:00:00 2001
From: Alex Deucher <[email protected]>
Date: Wed, 2 Dec 2009 11:46:52 -0500
Subject: drm/radeon/kms: handle vblanks properly with dpms on

From: Alex Deucher <[email protected]>

commit 500b758725314ab1b5316eb0caa5b0fa26740e6b upstream.

avivo chips

Copied from pre-avivo code.

Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>

---
drivers/gpu/drm/radeon/atombios_crtc.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -241,6 +241,7 @@ void atombios_crtc_dpms(struct drm_crtc
{
struct drm_device *dev = crtc->dev;
struct radeon_device *rdev = dev->dev_private;
+ struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);

switch (mode) {
case DRM_MODE_DPMS_ON:
@@ -248,20 +249,19 @@ void atombios_crtc_dpms(struct drm_crtc
if (ASIC_IS_DCE3(rdev))
atombios_enable_crtc_memreq(crtc, 1);
atombios_blank_crtc(crtc, 0);
+ drm_vblank_post_modeset(dev, radeon_crtc->crtc_id);
+ radeon_crtc_load_lut(crtc);
break;
case DRM_MODE_DPMS_STANDBY:
case DRM_MODE_DPMS_SUSPEND:
case DRM_MODE_DPMS_OFF:
+ drm_vblank_pre_modeset(dev, radeon_crtc->crtc_id);
atombios_blank_crtc(crtc, 1);
if (ASIC_IS_DCE3(rdev))
atombios_enable_crtc_memreq(crtc, 0);
atombios_enable_crtc(crtc, 0);
break;
}
-
- if (mode != DRM_MODE_DPMS_OFF) {
- radeon_crtc_load_lut(crtc);
- }
}

static void


After reverting this patch, modesetting works again. It's a Radeon HD3650 AGP
(RV635) on a Tyan Tiger K8W S2875ANRF (AMD 8151 AGP tunnel). Display is an
Acer X223W connected on DVI.


regards,
Johannes


2009-12-20 20:39:52

by Michael Tokarev

[permalink] [raw]
Subject: Re: [stable] Regression in 2.6.32.2: radeon KMS hangs system

Johannes Hirte wrote:
> With 2.6.32.2 the system hangs completely with black screen when modprobing
> radeon with modeset=1. I was able to connect via ssh, but after entering the
> password nothing happend. I didn't got a prompt. Only SysRq worked. I've
> tracked it down to the following patch:
>
> From 500b758725314ab1b5316eb0caa5b0fa26740e6b Mon Sep 17 00:00:00 2001
> From: Alex Deucher <[email protected]>
> Date: Wed, 2 Dec 2009 11:46:52 -0500
> Subject: drm/radeon/kms: handle vblanks properly with dpms on

Confirmed here on amd780g chipset. I was about to debug/bisect the issue
right when the email arrived. Reverting this patch makes my system bootable
again.

/mjt

> From: Alex Deucher <[email protected]>
>
> commit 500b758725314ab1b5316eb0caa5b0fa26740e6b upstream.
>
> avivo chips
>
> Copied from pre-avivo code.
>
> Signed-off-by: Alex Deucher <[email protected]>
> Signed-off-by: Dave Airlie <[email protected]>
> Signed-off-by: Greg Kroah-Hartman <[email protected]>
...

2009-12-21 16:13:59

by Alex Deucher

[permalink] [raw]
Subject: Re: [stable] Regression in 2.6.32.2: radeon KMS hangs system

On Sun, Dec 20, 2009 at 11:23 AM, Johannes Hirte
<[email protected]> wrote:
> With 2.6.32.2 the system hangs completely with black screen when modprobing
> radeon with modeset=1. I was able to connect via ssh, but after entering the
> password nothing happend. I didn't got a prompt. Only SysRq worked. I've
> tracked it down to the following patch:
>

Fixed here:
http://marc.info/?l=dri-devel&m=126137027403059&w=2

2.6.32 didn't have irq support for r6xx+.

Alex

> >From 500b758725314ab1b5316eb0caa5b0fa26740e6b Mon Sep 17 00:00:00 2001
> From: Alex Deucher <[email protected]>
> Date: Wed, 2 Dec 2009 11:46:52 -0500
> Subject: drm/radeon/kms: handle vblanks properly with dpms on
>
> From: Alex Deucher <[email protected]>
>
> commit 500b758725314ab1b5316eb0caa5b0fa26740e6b upstream.
>
> avivo chips
>
> Copied from pre-avivo code.
>
> Signed-off-by: Alex Deucher <[email protected]>
> Signed-off-by: Dave Airlie <[email protected]>
> Signed-off-by: Greg Kroah-Hartman <[email protected]>
>
> ---
> ?drivers/gpu/drm/radeon/atombios_crtc.c | ? ?8 ++++----
> ?1 file changed, 4 insertions(+), 4 deletions(-)
>
> --- a/drivers/gpu/drm/radeon/atombios_crtc.c
> +++ b/drivers/gpu/drm/radeon/atombios_crtc.c
> @@ -241,6 +241,7 @@ void atombios_crtc_dpms(struct drm_crtc
> ?{
> ? ? ? ?struct drm_device *dev = crtc->dev;
> ? ? ? ?struct radeon_device *rdev = dev->dev_private;
> + ? ? ? struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
>
> ? ? ? ?switch (mode) {
> ? ? ? ?case DRM_MODE_DPMS_ON:
> @@ -248,20 +249,19 @@ void atombios_crtc_dpms(struct drm_crtc
> ? ? ? ? ? ? ? ?if (ASIC_IS_DCE3(rdev))
> ? ? ? ? ? ? ? ? ? ? ? ?atombios_enable_crtc_memreq(crtc, 1);
> ? ? ? ? ? ? ? ?atombios_blank_crtc(crtc, 0);
> + ? ? ? ? ? ? ? drm_vblank_post_modeset(dev, radeon_crtc->crtc_id);
> + ? ? ? ? ? ? ? radeon_crtc_load_lut(crtc);
> ? ? ? ? ? ? ? ?break;
> ? ? ? ?case DRM_MODE_DPMS_STANDBY:
> ? ? ? ?case DRM_MODE_DPMS_SUSPEND:
> ? ? ? ?case DRM_MODE_DPMS_OFF:
> + ? ? ? ? ? ? ? drm_vblank_pre_modeset(dev, radeon_crtc->crtc_id);
> ? ? ? ? ? ? ? ?atombios_blank_crtc(crtc, 1);
> ? ? ? ? ? ? ? ?if (ASIC_IS_DCE3(rdev))
> ? ? ? ? ? ? ? ? ? ? ? ?atombios_enable_crtc_memreq(crtc, 0);
> ? ? ? ? ? ? ? ?atombios_enable_crtc(crtc, 0);
> ? ? ? ? ? ? ? ?break;
> ? ? ? ?}
> -
> - ? ? ? if (mode != DRM_MODE_DPMS_OFF) {
> - ? ? ? ? ? ? ? radeon_crtc_load_lut(crtc);
> - ? ? ? }
> ?}
>
> ?static void
>
>
> After reverting this patch, modesetting works again. It's a Radeon HD3650 AGP
> (RV635) on a Tyan Tiger K8W S2875ANRF (AMD 8151 AGP tunnel). Display is an
> Acer X223W connected on DVI.
>
>
> regards,
> ?Johannes
>
> ------------------------------------------------------------------------------
> This SF.Net email is sponsored by the Verizon Developer Community
> Take advantage of Verizon's best-in-class app development support
> A streamlined, 14 day to market process makes app distribution fast and easy
> Join now and get one step closer to millions of Verizon customers
> http://p.sf.net/sfu/verizon-dev2dev
> --
> _______________________________________________
> Dri-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/dri-devel
>