[REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode

public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed

* [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
@ 2018-05-21 19:53 Vito Caputo
  2018-05-21 21:57 ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-21 19:53 UTC (permalink / raw)
  To: linux-kernel

[-- Attachment #1: Type: text/plain, Size: 1841 bytes --]

Hello all,

4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
without -window.  I will try find time to build the latest rc this
evening.

> ~$ xgalaga
> X Error of failed request:  BadValue (integer parameter out of range for operation)
>   Major opcode of failed request:  152 (XFree86-VidModeExtension)
>   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
>   Value in failed request:  0x120004e
>   Serial number of failed request:  199
>   Current serial number in output stream:  203

Haven't dug into this much yet, only did a perfunctory check by booting into a
few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
It appears to be a 4.17-specific regression right now.

Also observed, though this is surely a different regression, the game
ran like molasses with -window, showing some prominent kworkers in top:

  692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
  102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
   94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
  696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
  902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
  891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
  903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss

The windowed performance issue was observed on the older kernels tested
as well, though 4.11 felt better and didn't have the busy kworkers.

I have not attempted to play xgalaga for ages, but it used to be perfectly
playable on this machine in windowed mode when I last did.

Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.

Regards,
Vito Caputo

[-- Attachment #2: 4.17-rc4-config.gz --]
[-- Type: application/gzip, Size: 24464 bytes --]

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-21 19:53 [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode Vito Caputo
@ 2018-05-21 21:57 ` Vito Caputo
  2018-05-23  9:49   ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-21 21:57 UTC (permalink / raw)
  To: linux-kernel

On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> Hello all,
> 
> 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> without -window.  I will try find time to build the latest rc this
> evening.
> 
> > ~$ xgalaga
> > X Error of failed request:  BadValue (integer parameter out of range for operation)
> >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> >   Value in failed request:  0x120004e
> >   Serial number of failed request:  199
> >   Current serial number in output stream:  203
> 
> Haven't dug into this much yet, only did a perfunctory check by booting into a
> few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> It appears to be a 4.17-specific regression right now.
> 
> Also observed, though this is surely a different regression, the game
> ran like molasses with -window, showing some prominent kworkers in top:
> 
>   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
>   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
>    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
>   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
>   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
>   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
>   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> 
> The windowed performance issue was observed on the older kernels tested
> as well, though 4.11 felt better and didn't have the busy kworkers.
> 
> I have not attempted to play xgalaga for ages, but it used to be perfectly
> playable on this machine in windowed mode when I last did.
> 
> Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> 

Just built and booted v4.17-rc6, still broken.

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-21 21:57 ` Vito Caputo
@ 2018-05-23  9:49   ` Vito Caputo
  2018-05-23 13:18     ` Ville Syrjälä
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-23  9:49 UTC (permalink / raw)
  To: linux-kernel; +Cc: ville.syrjala, daniel.vetter

On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > Hello all,
> > 
> > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > without -window.  I will try find time to build the latest rc this
> > evening.
> > 
> > > ~$ xgalaga
> > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > >   Value in failed request:  0x120004e
> > >   Serial number of failed request:  199
> > >   Current serial number in output stream:  203
> > 
> > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > It appears to be a 4.17-specific regression right now.
> > 
> > Also observed, though this is surely a different regression, the game
> > ran like molasses with -window, showing some prominent kworkers in top:
> > 
> >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > 
> > The windowed performance issue was observed on the older kernels tested
> > as well, though 4.11 felt better and didn't have the busy kworkers.
> > 
> > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > playable on this machine in windowed mode when I last did.
> > 
> > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > 
> 
> Just built and booted v4.17-rc6, still broken.

Bisected to:

e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
commit e995ca0b8139c5f6807095464e969931b443f55a
Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
Date:   Tue Nov 14 20:32:58 2017 +0200

    drm/i915: Provide a device level .mode_valid() hook
    
    We never support certain mode flags etc. Reject those early on in the
    mode_config.mode_valid() hook. That allows us to remove some duplicated
    checks from the connector .mode_valid() hooks, and it guarantees that
    we never see those flags even from user mode as the
    mode_config.mode_valid() hooks gets executed for those as well.
    
    Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
    Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
    Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Will CC Ville and Daniel.

Regards,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23  9:49   ` Vito Caputo
@ 2018-05-23 13:18     ` Ville Syrjälä
  2018-05-23 18:06       ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Ville Syrjälä @ 2018-05-23 13:18 UTC (permalink / raw)
  To: Vito Caputo; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > Hello all,
> > > 
> > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > without -window.  I will try find time to build the latest rc this
> > > evening.
> > > 
> > > > ~$ xgalaga
> > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > >   Value in failed request:  0x120004e
> > > >   Serial number of failed request:  199
> > > >   Current serial number in output stream:  203
> > > 
> > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > It appears to be a 4.17-specific regression right now.
> > > 
> > > Also observed, though this is surely a different regression, the game
> > > ran like molasses with -window, showing some prominent kworkers in top:
> > > 
> > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > 
> > > The windowed performance issue was observed on the older kernels tested
> > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > 
> > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > playable on this machine in windowed mode when I last did.
> > > 
> > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > 
> > 
> > Just built and booted v4.17-rc6, still broken.
> 
> Bisected to:
> 
> e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> commit e995ca0b8139c5f6807095464e969931b443f55a
> Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> Date:   Tue Nov 14 20:32:58 2017 +0200
> 
>     drm/i915: Provide a device level .mode_valid() hook
>     
>     We never support certain mode flags etc. Reject those early on in the
>     mode_config.mode_valid() hook. That allows us to remove some duplicated
>     checks from the connector .mode_valid() hooks, and it guarantees that
>     we never see those flags even from user mode as the
>     mode_config.mode_valid() hooks gets executed for those as well.
>     
>     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
>     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
>     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
the ddx doesn't validate before passing it on to the kernel. So far
I can't reproduce the problem here unfortnately.

Can you try the following patch and reproduce the problem with
drm.debug=0xe passed to the kernel so that we can seewhat the bad
modeline looks like?

>From c8b4eaaf3ee8e796dbbb4a5080c1aec2b435d110 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Wed, 23 May 2018 16:12:08 +0300
Subject: [PATCH] drm/modes: Dump the rejected user mode
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Print put the modeline When we reject a user mode.

Cc: Vito Caputo <vcaputo@pengaru.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 drivers/gpu/drm/drm_modes.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c
index c78ca0e84ffd..ca50b12d9701 100644
--- a/drivers/gpu/drm/drm_modes.c
+++ b/drivers/gpu/drm/drm_modes.c
@@ -1739,8 +1739,11 @@ int drm_mode_convert_umode(struct drm_device *dev,
 	}
 
 	out->status = drm_mode_validate_driver(dev, out);
-	if (out->status != MODE_OK)
+	if (out->status != MODE_OK) {
+		DRM_DEBUG_KMS("Bad user mode:\n");
+		drm_mode_debug_printmodeline(out);
 		return -EINVAL;
+	}
 
 	drm_mode_set_crtcinfo(out, CRTC_INTERLACE_HALVE_V);
 
-- 
2.16.1

-- 
Ville Syrjälä
Intel

^ permalink raw reply related	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 13:18     ` Ville Syrjälä
@ 2018-05-23 18:06       ` Vito Caputo
  2018-05-23 18:20         ` Ville Syrjälä
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-23 18:06 UTC (permalink / raw)
  To: Ville Syrjälä; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > Hello all,
> > > > 
> > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > without -window.  I will try find time to build the latest rc this
> > > > evening.
> > > > 
> > > > > ~$ xgalaga
> > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > >   Value in failed request:  0x120004e
> > > > >   Serial number of failed request:  199
> > > > >   Current serial number in output stream:  203
> > > > 
> > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > It appears to be a 4.17-specific regression right now.
> > > > 
> > > > Also observed, though this is surely a different regression, the game
> > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > 
> > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > 
> > > > The windowed performance issue was observed on the older kernels tested
> > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > 
> > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > playable on this machine in windowed mode when I last did.
> > > > 
> > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > 
> > > 
> > > Just built and booted v4.17-rc6, still broken.
> > 
> > Bisected to:
> > 
> > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > commit e995ca0b8139c5f6807095464e969931b443f55a
> > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > Date:   Tue Nov 14 20:32:58 2017 +0200
> > 
> >     drm/i915: Provide a device level .mode_valid() hook
> >     
> >     We never support certain mode flags etc. Reject those early on in the
> >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> >     checks from the connector .mode_valid() hooks, and it guarantees that
> >     we never see those flags even from user mode as the
> >     mode_config.mode_valid() hooks gets executed for those as well.
> >     
> >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> 
> Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> the ddx doesn't validate before passing it on to the kernel. So far
> I can't reproduce the problem here unfortnately.
> 
> Can you try the following patch and reproduce the problem with
> drm.debug=0xe passed to the kernel so that we can seewhat the bad
> modeline looks like?
> 

dmesg after xgalaga fails:

```
[   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
[   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
[   75.617458] [drm:drm_mode_setcrtc] Invalid mode
```

xrandr --verbose:

```
Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
	Identifier: 0x41
	Timestamp:  23375
	Subpixel:   horizontal rgb
	Gamma:      1.0:1.0:1.0
	Brightness: 1.0
	Clones:    
	CRTC:       0
	CRTCs:      0 1
	Transform:  1.000000 0.000000 0.000000
	            0.000000 1.000000 0.000000
	            0.000000 0.000000 1.000000
	           filter: 
	EDID: 
		00ffffffffffff0030ae004000000000
		3010010380191278eafe609555518726
		22505421080001010101010101010101
		01010101010128150040410026301888
		3600f6b800000018ed10004041002630
		18883600f6b9000000180000000f0061
		43326143280f01000daf0714000000fe
		004e31323158352d4c303620202000ed
	scaling mode: Full aspect 
		supported: Full, Center, Full aspect
	non-desktop: 0 
		range: (0, 1)
	link-status: Good 
		supported: Good, Bad
  1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
        v: height  768 start  771 end  777 total  806           clock  50.00Hz
  1024x768 (0x45) 65.000MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
        v: height  768 start  771 end  777 total  806           clock  60.00Hz
  1024x768 (0x46) 43.330MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
        v: height  768 start  771 end  777 total  806           clock  40.00Hz
  960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
        h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
        v: height  720 start  720 end  722 total  750           clock  60.00Hz
  928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
        h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
        v: height  696 start  696 end  698 total  719           clock  60.05Hz
  896x672 (0x49) 102.400MHz -HSync +VSync DoubleScan
        h: width   896 start  960 end 1060 total 1224 skew    0 clock  83.66KHz
        v: height  672 start  672 end  674 total  697           clock  60.01Hz
  960x600 (0x4a) 77.000MHz +HSync -VSync DoubleScan
        h: width   960 start  984 end 1000 total 1040 skew    0 clock  74.04KHz
        v: height  600 start  601 end  604 total  617           clock  60.00Hz
  960x540 (0x4b) 69.250MHz +HSync -VSync DoubleScan
        h: width   960 start  984 end 1000 total 1040 skew    0 clock  66.59KHz
        v: height  540 start  541 end  544 total  555           clock  59.99Hz
  800x600 (0x4c) 81.000MHz +HSync +VSync DoubleScan
        h: width   800 start  832 end  928 total 1080 skew    0 clock  75.00KHz
        v: height  600 start  600 end  602 total  625           clock  60.00Hz
  800x600 (0x4d) 40.000MHz +HSync +VSync
        h: width   800 start  840 end  968 total 1056 skew    0 clock  37.88KHz
        v: height  600 start  601 end  605 total  628           clock  60.32Hz
  800x600 (0x4e) 36.000MHz +HSync +VSync
        h: width   800 start  824 end  896 total 1024 skew    0 clock  35.16KHz
        v: height  600 start  601 end  603 total  625           clock  56.25Hz
  840x525 (0x4f) 73.125MHz -HSync +VSync DoubleScan
        h: width   840 start  892 end  980 total 1120 skew    0 clock  65.29KHz
        v: height  525 start  526 end  529 total  544           clock  60.01Hz
  840x525 (0x50) 59.500MHz +HSync -VSync DoubleScan
        h: width   840 start  864 end  880 total  920 skew    0 clock  64.67KHz
        v: height  525 start  526 end  529 total  540           clock  59.88Hz
  800x512 (0x51) 51.562MHz +HSync +VSync DoubleScan
        h: width   800 start  800 end  828 total  832 skew    0 clock  61.97KHz
        v: height  512 start  512 end  514 total  515           clock  60.17Hz
  700x525 (0x52) 61.000MHz +HSync +VSync DoubleScan
        h: width   700 start  744 end  820 total  940 skew    0 clock  64.89KHz
        v: height  525 start  526 end  532 total  541           clock  59.98Hz
  640x512 (0x53) 54.000MHz +HSync +VSync DoubleScan
        h: width   640 start  664 end  720 total  844 skew    0 clock  63.98KHz
        v: height  512 start  512 end  514 total  533           clock  60.02Hz
  720x450 (0x54) 53.250MHz -HSync +VSync DoubleScan
        h: width   720 start  760 end  836 total  952 skew    0 clock  55.93KHz
        v: height  450 start  451 end  454 total  467           clock  59.89Hz
  640x480 (0x55) 54.000MHz +HSync +VSync DoubleScan
        h: width   640 start  688 end  744 total  900 skew    0 clock  60.00KHz
        v: height  480 start  480 end  482 total  500           clock  60.00Hz
  640x480 (0x56) 25.175MHz -HSync -VSync
        h: width   640 start  656 end  752 total  800 skew    0 clock  31.47KHz
        v: height  480 start  490 end  492 total  525           clock  59.94Hz
  680x384 (0x57) 42.375MHz -HSync +VSync DoubleScan
        h: width   680 start  716 end  784 total  888 skew    0 clock  47.72KHz
        v: height  384 start  385 end  390 total  399           clock  59.80Hz
  680x384 (0x58) 36.000MHz +HSync -VSync DoubleScan
        h: width   680 start  704 end  720 total  760 skew    0 clock  47.37KHz
        v: height  384 start  385 end  390 total  395           clock  59.96Hz
  576x432 (0x59) 40.810MHz -HSync +VSync DoubleScan
        h: width   576 start  608 end  668 total  760 skew    0 clock  53.70KHz
        v: height  432 start  432 end  434 total  447           clock  60.06Hz
  512x384 (0x5a) 32.500MHz -HSync -VSync DoubleScan
        h: width   512 start  524 end  592 total  672 skew    0 clock  48.36KHz
        v: height  384 start  385 end  388 total  403           clock  60.00Hz
  400x300 (0x5b) 20.000MHz +HSync +VSync DoubleScan
        h: width   400 start  420 end  484 total  528 skew    0 clock  37.88KHz
        v: height  300 start  300 end  302 total  314           clock  60.32Hz
  400x300 (0x5c) 18.000MHz +HSync +VSync DoubleScan
        h: width   400 start  412 end  448 total  512 skew    0 clock  35.16KHz
        v: height  300 start  300 end  301 total  312           clock  56.34Hz
  320x240 (0x5d) 12.587MHz -HSync -VSync DoubleScan
        h: width   320 start  328 end  376 total  400 skew    0 clock  31.47KHz
        v: height  240 start  245 end  246 total  262           clock  60.05Hz
VGA-1 disconnected (normal left inverted right x axis y axis)
	Identifier: 0x42
	Timestamp:  23375
	Subpixel:   unknown
	Clones:    
	CRTCs:      0 1
	Transform:  1.000000 0.000000 0.000000
	            0.000000 1.000000 0.000000
	            0.000000 0.000000 1.000000
	           filter: 
	non-desktop: 0 
		range: (0, 1)
	link-status: Good 
		supported: Good, Bad
```

I took a quick glance at the xgalaga source, and it looks to supply
XF86VidModeSwitchToMode() with an unmodified XF86VidModeModeInfo chosen from
the array returned by XF86VidModeGetAllModeLines().

Thanks,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 18:06       ` Vito Caputo
@ 2018-05-23 18:20         ` Ville Syrjälä
  2018-05-23 18:39           ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Ville Syrjälä @ 2018-05-23 18:20 UTC (permalink / raw)
  To: Vito Caputo; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > Hello all,
> > > > > 
> > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > without -window.  I will try find time to build the latest rc this
> > > > > evening.
> > > > > 
> > > > > > ~$ xgalaga
> > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > >   Value in failed request:  0x120004e
> > > > > >   Serial number of failed request:  199
> > > > > >   Current serial number in output stream:  203
> > > > > 
> > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > It appears to be a 4.17-specific regression right now.
> > > > > 
> > > > > Also observed, though this is surely a different regression, the game
> > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > 
> > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > 
> > > > > The windowed performance issue was observed on the older kernels tested
> > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > 
> > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > playable on this machine in windowed mode when I last did.
> > > > > 
> > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > 
> > > > 
> > > > Just built and booted v4.17-rc6, still broken.
> > > 
> > > Bisected to:
> > > 
> > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > 
> > >     drm/i915: Provide a device level .mode_valid() hook
> > >     
> > >     We never support certain mode flags etc. Reject those early on in the
> > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > >     we never see those flags even from user mode as the
> > >     mode_config.mode_valid() hooks gets executed for those as well.
> > >     
> > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > 
> > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > the ddx doesn't validate before passing it on to the kernel. So far
> > I can't reproduce the problem here unfortnately.
> > 
> > Can you try the following patch and reproduce the problem with
> > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > modeline looks like?
> > 
> 
> dmesg after xgalaga fails:
> 
> ```
> [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25

0x20 == dblscan

> [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> ```
> 
> xrandr --verbose:
> 
> ```
> Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> 	Identifier: 0x41
> 	Timestamp:  23375
> 	Subpixel:   horizontal rgb
> 	Gamma:      1.0:1.0:1.0
> 	Brightness: 1.0
> 	Clones:    
> 	CRTC:       0
> 	CRTCs:      0 1
> 	Transform:  1.000000 0.000000 0.000000
> 	            0.000000 1.000000 0.000000
> 	            0.000000 0.000000 1.000000
> 	           filter: 
> 	EDID: 
> 		00ffffffffffff0030ae004000000000
> 		3010010380191278eafe609555518726
> 		22505421080001010101010101010101
> 		01010101010128150040410026301888
> 		3600f6b800000018ed10004041002630
> 		18883600f6b9000000180000000f0061
> 		43326143280f01000daf0714000000fe
> 		004e31323158352d4c303620202000ed
> 	scaling mode: Full aspect 
> 		supported: Full, Center, Full aspect
> 	non-desktop: 0 
> 		range: (0, 1)
> 	link-status: Good 
> 		supported: Good, Bad
>   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
>         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
>         v: height  768 start  771 end  777 total  806           clock  50.00Hz
>   1024x768 (0x45) 65.000MHz -HSync -VSync
>         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
>         v: height  768 start  771 end  777 total  806           clock  60.00Hz
>   1024x768 (0x46) 43.330MHz -HSync -VSync
>         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
>         v: height  768 start  771 end  777 total  806           clock  40.00Hz
>   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
>         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
>         v: height  720 start  720 end  722 total  750           clock  60.00Hz
>   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
>         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
>         v: height  696 start  696 end  698 total  719           clock  60.05Hz

Where are all these dblscan modes coming from?

Did you add them manually or are they being automatically
generated by something?

-- 
Ville Syrjälä
Intel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 18:20         ` Ville Syrjälä
@ 2018-05-23 18:39           ` Vito Caputo
  2018-05-23 19:12             ` Ville Syrjälä
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-23 18:39 UTC (permalink / raw)
  To: Ville Syrjälä; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > Hello all,
> > > > > > 
> > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > evening.
> > > > > > 
> > > > > > > ~$ xgalaga
> > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > >   Value in failed request:  0x120004e
> > > > > > >   Serial number of failed request:  199
> > > > > > >   Current serial number in output stream:  203
> > > > > > 
> > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > 
> > > > > > Also observed, though this is surely a different regression, the game
> > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > 
> > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > 
> > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > 
> > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > playable on this machine in windowed mode when I last did.
> > > > > > 
> > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > 
> > > > > 
> > > > > Just built and booted v4.17-rc6, still broken.
> > > > 
> > > > Bisected to:
> > > > 
> > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > 
> > > >     drm/i915: Provide a device level .mode_valid() hook
> > > >     
> > > >     We never support certain mode flags etc. Reject those early on in the
> > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > >     we never see those flags even from user mode as the
> > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > >     
> > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > 
> > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > the ddx doesn't validate before passing it on to the kernel. So far
> > > I can't reproduce the problem here unfortnately.
> > > 
> > > Can you try the following patch and reproduce the problem with
> > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > modeline looks like?
> > > 
> > 
> > dmesg after xgalaga fails:
> > 
> > ```
> > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> 
> 0x20 == dblscan
> 
> > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > ```
> > 
> > xrandr --verbose:
> > 
> > ```
> > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > 	Identifier: 0x41
> > 	Timestamp:  23375
> > 	Subpixel:   horizontal rgb
> > 	Gamma:      1.0:1.0:1.0
> > 	Brightness: 1.0
> > 	Clones:    
> > 	CRTC:       0
> > 	CRTCs:      0 1
> > 	Transform:  1.000000 0.000000 0.000000
> > 	            0.000000 1.000000 0.000000
> > 	            0.000000 0.000000 1.000000
> > 	           filter: 
> > 	EDID: 
> > 		00ffffffffffff0030ae004000000000
> > 		3010010380191278eafe609555518726
> > 		22505421080001010101010101010101
> > 		01010101010128150040410026301888
> > 		3600f6b800000018ed10004041002630
> > 		18883600f6b9000000180000000f0061
> > 		43326143280f01000daf0714000000fe
> > 		004e31323158352d4c303620202000ed
> > 	scaling mode: Full aspect 
> > 		supported: Full, Center, Full aspect
> > 	non-desktop: 0 
> > 		range: (0, 1)
> > 	link-status: Good 
> > 		supported: Good, Bad
> >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> >   1024x768 (0x45) 65.000MHz -HSync -VSync
> >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> >   1024x768 (0x46) 43.330MHz -HSync -VSync
> >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> 
> Where are all these dblscan modes coming from?
> 
> Did you add them manually or are they being automatically
> generated by something?
> 

This is pure automagic xserver-xorg-video-intel on i915 drm/kms.

After booting back into a working 4.16 kernel I see the same xrandr --verbose
list with all the DoubleScan modes, with which xgalaga is functional.

There's this in the Xorg.log:

```
[    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
[    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
[    12.882] (II) intel(0): Output VGA1 has no monitor section
[    12.883] (II) intel(0): EDID for output LVDS1
[    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
[    12.883] (II) intel(0): Year: 2006  Week: 48
[    12.883] (II) intel(0): EDID Version: 1.3
[    12.883] (II) intel(0): Digital Display Input
[    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
[    12.883] (II) intel(0): Gamma: 2.20
[    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
[    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
[    12.883] (II) intel(0): First detailed timing is preferred mode
[    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
[    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
[    12.883] (II) intel(0): Supported established timings:
[    12.883] (II) intel(0): 640x480@60Hz
[    12.883] (II) intel(0): 800x600@60Hz
[    12.883] (II) intel(0): 1024x768@60Hz
[    12.883] (II) intel(0): Manufacturer's mask: 0
[    12.883] (II) intel(0): Supported detailed timing:
[    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
[    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
[    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
[    12.883] (II) intel(0): Supported detailed timing:
[    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
[    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
[    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
[    12.883] (II) intel(0): Unknown vendor-specific block f
[    12.883] (II) intel(0):  N121X5-L06
[    12.883] (II) intel(0): EDID (in hex):
[    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
[    12.883] (II) intel(0):     3010010380191278eafe609555518726
[    12.883] (II) intel(0):     22505421080001010101010101010101
[    12.883] (II) intel(0):     01010101010128150040410026301888
[    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
[    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
[    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
[    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
[    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
[    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
[    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
[    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)
[    12.884] (II) intel(0): Printing probed modes for output LVDS1
[    12.884] (II) intel(0): Modeline "1024x768"x50.0   54.16  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (40.3 kHz eP)
[    12.884] (II) intel(0): Modeline "1024x768"x60.0   65.00  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (48.4 kHz e)
[    12.884] (II) intel(0): Modeline "1024x768"x40.0   43.33  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (32.2 kHz e)
[    12.884] (II) intel(0): Modeline "800x600"x60.3   40.00  800 840 968 1056  600 601 605 628 +hsync +vsync (37.9 kHz e)
[    12.884] (II) intel(0): Modeline "800x600"x56.2   36.00  800 824 896 1024  600 601 603 625 +hsync +vsync (35.2 kHz d)
[    12.884] (II) intel(0): Modeline "640x480"x59.9   25.18  640 656 752 800  480 490 492 525 -hsync -vsync (31.5 kHz e)
[    12.906] (II) intel(0): EDID for output VGA1
[    12.906] (II) intel(0): Output LVDS1 connected
[    12.906] (II) intel(0): Output VGA1 disconnected
[    12.906] (II) intel(0): Using exact sizes for initial modes
```

Thanks,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 18:39           ` Vito Caputo
@ 2018-05-23 19:12             ` Ville Syrjälä
  2018-05-23 20:15               ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Ville Syrjälä @ 2018-05-23 19:12 UTC (permalink / raw)
  To: Vito Caputo; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 11:39:00AM -0700, Vito Caputo wrote:
> On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> > On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > > Hello all,
> > > > > > > 
> > > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > > evening.
> > > > > > > 
> > > > > > > > ~$ xgalaga
> > > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > > >   Value in failed request:  0x120004e
> > > > > > > >   Serial number of failed request:  199
> > > > > > > >   Current serial number in output stream:  203
> > > > > > > 
> > > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > > 
> > > > > > > Also observed, though this is surely a different regression, the game
> > > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > > 
> > > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > > 
> > > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > > 
> > > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > > playable on this machine in windowed mode when I last did.
> > > > > > > 
> > > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > > 
> > > > > > 
> > > > > > Just built and booted v4.17-rc6, still broken.
> > > > > 
> > > > > Bisected to:
> > > > > 
> > > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > > 
> > > > >     drm/i915: Provide a device level .mode_valid() hook
> > > > >     
> > > > >     We never support certain mode flags etc. Reject those early on in the
> > > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > > >     we never see those flags even from user mode as the
> > > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > > >     
> > > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > 
> > > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > > the ddx doesn't validate before passing it on to the kernel. So far
> > > > I can't reproduce the problem here unfortnately.
> > > > 
> > > > Can you try the following patch and reproduce the problem with
> > > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > > modeline looks like?
> > > > 
> > > 
> > > dmesg after xgalaga fails:
> > > 
> > > ```
> > > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> > 
> > 0x20 == dblscan
> > 
> > > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > > ```
> > > 
> > > xrandr --verbose:
> > > 
> > > ```
> > > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > > 	Identifier: 0x41
> > > 	Timestamp:  23375
> > > 	Subpixel:   horizontal rgb
> > > 	Gamma:      1.0:1.0:1.0
> > > 	Brightness: 1.0
> > > 	Clones:    
> > > 	CRTC:       0
> > > 	CRTCs:      0 1
> > > 	Transform:  1.000000 0.000000 0.000000
> > > 	            0.000000 1.000000 0.000000
> > > 	            0.000000 0.000000 1.000000
> > > 	           filter: 
> > > 	EDID: 
> > > 		00ffffffffffff0030ae004000000000
> > > 		3010010380191278eafe609555518726
> > > 		22505421080001010101010101010101
> > > 		01010101010128150040410026301888
> > > 		3600f6b800000018ed10004041002630
> > > 		18883600f6b9000000180000000f0061
> > > 		43326143280f01000daf0714000000fe
> > > 		004e31323158352d4c303620202000ed
> > > 	scaling mode: Full aspect 
> > > 		supported: Full, Center, Full aspect
> > > 	non-desktop: 0 
> > > 		range: (0, 1)
> > > 	link-status: Good 
> > > 		supported: Good, Bad
> > >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> > >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> > >   1024x768 (0x45) 65.000MHz -HSync -VSync
> > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> > >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> > >   1024x768 (0x46) 43.330MHz -HSync -VSync
> > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> > >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> > >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> > >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> > >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> > >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> > >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> > >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> > 
> > Where are all these dblscan modes coming from?
> > 
> > Did you add them manually or are they being automatically
> > generated by something?
> > 
> 
> This is pure automagic xserver-xorg-video-intel on i915 drm/kms.
> 
> After booting back into a working 4.16 kernel I see the same xrandr --verbose
> list with all the DoubleScan modes, with which xgalaga is functional.
> 
> There's this in the Xorg.log:
> 
> ```
> [    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
> [    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
> [    12.882] (II) intel(0): Output VGA1 has no monitor section
> [    12.883] (II) intel(0): EDID for output LVDS1
> [    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
> [    12.883] (II) intel(0): Year: 2006  Week: 48
> [    12.883] (II) intel(0): EDID Version: 1.3
> [    12.883] (II) intel(0): Digital Display Input
> [    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
> [    12.883] (II) intel(0): Gamma: 2.20
> [    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
> [    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
> [    12.883] (II) intel(0): First detailed timing is preferred mode
> [    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
> [    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
> [    12.883] (II) intel(0): Supported established timings:
> [    12.883] (II) intel(0): 640x480@60Hz
> [    12.883] (II) intel(0): 800x600@60Hz
> [    12.883] (II) intel(0): 1024x768@60Hz
> [    12.883] (II) intel(0): Manufacturer's mask: 0
> [    12.883] (II) intel(0): Supported detailed timing:
> [    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
> [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> [    12.883] (II) intel(0): Supported detailed timing:
> [    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
> [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> [    12.883] (II) intel(0): Unknown vendor-specific block f
> [    12.883] (II) intel(0):  N121X5-L06
> [    12.883] (II) intel(0): EDID (in hex):
> [    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
> [    12.883] (II) intel(0):     3010010380191278eafe609555518726
> [    12.883] (II) intel(0):     22505421080001010101010101010101
> [    12.883] (II) intel(0):     01010101010128150040410026301888
> [    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
> [    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
> [    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
> [    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
> [    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> [    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
> [    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)

So first it rejected them all, but somehow later they get added back?
Very odd.

Hmm. Must be coming from that default modes thing...

Looks pretty much like ddx bug to me. Does the following ddx patch get
rid of the bogus dblscan modes?

>From 40672b8527313d8bc01aa5cdcda8602687ec0f23 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Wed, 23 May 2018 22:10:07 +0300
Subject: [PATCH xf86-video-intel] Don't accept dblscan from xf86DefaultModes[]

---
 src/sna/sna_display.c   | 3 +++
 src/uxa/intel_display.c | 3 +++
 2 files changed, 6 insertions(+)

diff --git a/src/sna/sna_display.c b/src/sna/sna_display.c
index 6f6bea2e9cad..b38a915ca4ec 100644
--- a/src/sna/sna_display.c
+++ b/src/sna/sna_display.c
@@ -4204,6 +4204,8 @@ sna_output_add_default_modes(xf86OutputPtr output, DisplayModePtr modes)
 	DisplayModePtr i, m, preferred = NULL;
 	int max_x = 0, max_y = 0, max_clock = 0;
 	float max_vrefresh = 0.0;
+	int flags = (output->interlaceAllowed ? V_INTERLACE : 0) |
+		(output->doubleScanAllowed ? V_DBLSCAN : 0);
 
 	if (mon && GTF_SUPPORTED(mon->features.msc))
 		return modes;
@@ -4220,6 +4222,7 @@ sna_output_add_default_modes(xf86OutputPtr output, DisplayModePtr modes)
 
 	m = default_modes(preferred);
 	xf86ValidateModesSize(output->scrn, m, max_x, max_y, 0);
+	xf86ValidateModesFlags(output->scrn, m, flags);
 
 	for (i = m; i; i = i->next) {
 		if (i->Clock > max_clock)
diff --git a/src/uxa/intel_display.c b/src/uxa/intel_display.c
index 809cda1dd56b..431727fc4ac2 100644
--- a/src/uxa/intel_display.c
+++ b/src/uxa/intel_display.c
@@ -919,6 +919,8 @@ intel_output_panel_edid(xf86OutputPtr output, DisplayModePtr modes)
 		DisplayModePtr i, m, p = NULL;
 		int max_x = 0, max_y = 0;
 		float max_vrefresh = 0.0;
+		int flags = (output->interlaceAllowed ? V_INTERLACE : 0) |
+			(output->doubleScanAllowed ? V_DBLSCAN : 0);
 
 		for (m = modes; m; m = m->next) {
 			if (m->type & M_T_PREFERRED)
@@ -938,6 +940,7 @@ intel_output_panel_edid(xf86OutputPtr output, DisplayModePtr modes)
 #endif
 
 		xf86ValidateModesSize(output->scrn, m, max_x, max_y, 0);
+		xf86ValidateModesFlags(output->scrn, m, flags);
 
 		for (i = m; i; i = i->next) {
 			if (xf86ModeVRefresh(i) > max_vrefresh)
-- 
2.16.1

-- 
Ville Syrjälä
Intel

^ permalink raw reply related	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 19:12             ` Ville Syrjälä
@ 2018-05-23 20:15               ` Vito Caputo
  2018-05-23 20:41                 ` Ville Syrjälä
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-23 20:15 UTC (permalink / raw)
  To: Ville Syrjälä; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 10:12:22PM +0300, Ville Syrjälä wrote:
> On Wed, May 23, 2018 at 11:39:00AM -0700, Vito Caputo wrote:
> > On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> > > On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > > > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > > > Hello all,
> > > > > > > > 
> > > > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > > > evening.
> > > > > > > > 
> > > > > > > > > ~$ xgalaga
> > > > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > > > >   Value in failed request:  0x120004e
> > > > > > > > >   Serial number of failed request:  199
> > > > > > > > >   Current serial number in output stream:  203
> > > > > > > > 
> > > > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > > > 
> > > > > > > > Also observed, though this is surely a different regression, the game
> > > > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > > > 
> > > > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > > > 
> > > > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > > > 
> > > > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > > > playable on this machine in windowed mode when I last did.
> > > > > > > > 
> > > > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > > > 
> > > > > > > 
> > > > > > > Just built and booted v4.17-rc6, still broken.
> > > > > > 
> > > > > > Bisected to:
> > > > > > 
> > > > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > > > 
> > > > > >     drm/i915: Provide a device level .mode_valid() hook
> > > > > >     
> > > > > >     We never support certain mode flags etc. Reject those early on in the
> > > > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > > > >     we never see those flags even from user mode as the
> > > > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > > > >     
> > > > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > 
> > > > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > > > the ddx doesn't validate before passing it on to the kernel. So far
> > > > > I can't reproduce the problem here unfortnately.
> > > > > 
> > > > > Can you try the following patch and reproduce the problem with
> > > > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > > > modeline looks like?
> > > > > 
> > > > 
> > > > dmesg after xgalaga fails:
> > > > 
> > > > ```
> > > > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > > > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> > > 
> > > 0x20 == dblscan
> > > 
> > > > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > > > ```
> > > > 
> > > > xrandr --verbose:
> > > > 
> > > > ```
> > > > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > > > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > > > 	Identifier: 0x41
> > > > 	Timestamp:  23375
> > > > 	Subpixel:   horizontal rgb
> > > > 	Gamma:      1.0:1.0:1.0
> > > > 	Brightness: 1.0
> > > > 	Clones:    
> > > > 	CRTC:       0
> > > > 	CRTCs:      0 1
> > > > 	Transform:  1.000000 0.000000 0.000000
> > > > 	            0.000000 1.000000 0.000000
> > > > 	            0.000000 0.000000 1.000000
> > > > 	           filter: 
> > > > 	EDID: 
> > > > 		00ffffffffffff0030ae004000000000
> > > > 		3010010380191278eafe609555518726
> > > > 		22505421080001010101010101010101
> > > > 		01010101010128150040410026301888
> > > > 		3600f6b800000018ed10004041002630
> > > > 		18883600f6b9000000180000000f0061
> > > > 		43326143280f01000daf0714000000fe
> > > > 		004e31323158352d4c303620202000ed
> > > > 	scaling mode: Full aspect 
> > > > 		supported: Full, Center, Full aspect
> > > > 	non-desktop: 0 
> > > > 		range: (0, 1)
> > > > 	link-status: Good 
> > > > 		supported: Good, Bad
> > > >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> > > >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> > > >   1024x768 (0x45) 65.000MHz -HSync -VSync
> > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> > > >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> > > >   1024x768 (0x46) 43.330MHz -HSync -VSync
> > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> > > >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> > > >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> > > >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> > > >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> > > >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> > > >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> > > >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> > > 
> > > Where are all these dblscan modes coming from?
> > > 
> > > Did you add them manually or are they being automatically
> > > generated by something?
> > > 
> > 
> > This is pure automagic xserver-xorg-video-intel on i915 drm/kms.
> > 
> > After booting back into a working 4.16 kernel I see the same xrandr --verbose
> > list with all the DoubleScan modes, with which xgalaga is functional.
> > 
> > There's this in the Xorg.log:
> > 
> > ```
> > [    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
> > [    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
> > [    12.882] (II) intel(0): Output VGA1 has no monitor section
> > [    12.883] (II) intel(0): EDID for output LVDS1
> > [    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
> > [    12.883] (II) intel(0): Year: 2006  Week: 48
> > [    12.883] (II) intel(0): EDID Version: 1.3
> > [    12.883] (II) intel(0): Digital Display Input
> > [    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
> > [    12.883] (II) intel(0): Gamma: 2.20
> > [    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
> > [    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
> > [    12.883] (II) intel(0): First detailed timing is preferred mode
> > [    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
> > [    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
> > [    12.883] (II) intel(0): Supported established timings:
> > [    12.883] (II) intel(0): 640x480@60Hz
> > [    12.883] (II) intel(0): 800x600@60Hz
> > [    12.883] (II) intel(0): 1024x768@60Hz
> > [    12.883] (II) intel(0): Manufacturer's mask: 0
> > [    12.883] (II) intel(0): Supported detailed timing:
> > [    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
> > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > [    12.883] (II) intel(0): Supported detailed timing:
> > [    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
> > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > [    12.883] (II) intel(0): Unknown vendor-specific block f
> > [    12.883] (II) intel(0):  N121X5-L06
> > [    12.883] (II) intel(0): EDID (in hex):
> > [    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
> > [    12.883] (II) intel(0):     3010010380191278eafe609555518726
> > [    12.883] (II) intel(0):     22505421080001010101010101010101
> > [    12.883] (II) intel(0):     01010101010128150040410026301888
> > [    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
> > [    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
> > [    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
> > [    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
> > [    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > [    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
> > [    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)
> 
> So first it rejected them all, but somehow later they get added back?
> Very odd.
> 
> Hmm. Must be coming from that default modes thing...
> 
> Looks pretty much like ddx bug to me. Does the following ddx patch get
> rid of the bogus dblscan modes?
> 

<snip>

No, that didn't seem to change anything.  After some digging, I found:

xorg-server-1.19.2/hw/xfree86/drivers/modesetting/drmmode_display.c::drmmode_output_init():

```
1796     drmmode_output->output_id = mode_res->connectors[num];
1797     drmmode_output->mode_output = koutput;
1798     drmmode_output->mode_encoders = kencoders;
1799     drmmode_output->drmmode = drmmode;
1800     output->mm_width = koutput->mmWidth;
1801     output->mm_height = koutput->mmHeight;
1802 
1803     output->subpixel_order = subpixel_conv_table[koutput->subpixel];
1804     output->interlaceAllowed = TRUE;
1805     output->doubleScanAllowed = TRUE;
1806     output->driver_private = drmmode_output;
```

I don't see anywhere that sets interlaceAllowed or doubleScanAllowed to FALSE,
except xserver-xorg-video-intel/src/legacy/i810/i810_driver.c.

So your change never invalidates the doublescan modes, because they're allowed
for the output.

Assuming this can be corrected by disallowing these for the output
somewhere appropriate, isn't this still a userspace-breaking kernel
regression?

It seems to me like the kernel should just ignore the doublescan mode
flag and continue if a non-doublescan equivalent mode exists, given the
userspace situation this change uncovered.

Surely xgalaga isn't the only application using this API in such an
obvious fashion.

Thanks,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 20:15               ` Vito Caputo
@ 2018-05-23 20:41                 ` Ville Syrjälä
  2018-05-23 21:27                   ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Ville Syrjälä @ 2018-05-23 20:41 UTC (permalink / raw)
  To: Vito Caputo; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 01:15:05PM -0700, Vito Caputo wrote:
> On Wed, May 23, 2018 at 10:12:22PM +0300, Ville Syrjälä wrote:
> > On Wed, May 23, 2018 at 11:39:00AM -0700, Vito Caputo wrote:
> > > On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> > > > On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > > > > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > > > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > > > > Hello all,
> > > > > > > > > 
> > > > > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > > > > evening.
> > > > > > > > > 
> > > > > > > > > > ~$ xgalaga
> > > > > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > > > > >   Value in failed request:  0x120004e
> > > > > > > > > >   Serial number of failed request:  199
> > > > > > > > > >   Current serial number in output stream:  203
> > > > > > > > > 
> > > > > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > > > > 
> > > > > > > > > Also observed, though this is surely a different regression, the game
> > > > > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > > > > 
> > > > > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > > > > 
> > > > > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > > > > 
> > > > > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > > > > playable on this machine in windowed mode when I last did.
> > > > > > > > > 
> > > > > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > > > > 
> > > > > > > > 
> > > > > > > > Just built and booted v4.17-rc6, still broken.
> > > > > > > 
> > > > > > > Bisected to:
> > > > > > > 
> > > > > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > > > > 
> > > > > > >     drm/i915: Provide a device level .mode_valid() hook
> > > > > > >     
> > > > > > >     We never support certain mode flags etc. Reject those early on in the
> > > > > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > > > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > > > > >     we never see those flags even from user mode as the
> > > > > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > > > > >     
> > > > > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > > > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > > 
> > > > > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > > > > the ddx doesn't validate before passing it on to the kernel. So far
> > > > > > I can't reproduce the problem here unfortnately.
> > > > > > 
> > > > > > Can you try the following patch and reproduce the problem with
> > > > > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > > > > modeline looks like?
> > > > > > 
> > > > > 
> > > > > dmesg after xgalaga fails:
> > > > > 
> > > > > ```
> > > > > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > > > > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> > > > 
> > > > 0x20 == dblscan
> > > > 
> > > > > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > > > > ```
> > > > > 
> > > > > xrandr --verbose:
> > > > > 
> > > > > ```
> > > > > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > > > > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > > > > 	Identifier: 0x41
> > > > > 	Timestamp:  23375
> > > > > 	Subpixel:   horizontal rgb
> > > > > 	Gamma:      1.0:1.0:1.0
> > > > > 	Brightness: 1.0
> > > > > 	Clones:    
> > > > > 	CRTC:       0
> > > > > 	CRTCs:      0 1
> > > > > 	Transform:  1.000000 0.000000 0.000000
> > > > > 	            0.000000 1.000000 0.000000
> > > > > 	            0.000000 0.000000 1.000000
> > > > > 	           filter: 
> > > > > 	EDID: 
> > > > > 		00ffffffffffff0030ae004000000000
> > > > > 		3010010380191278eafe609555518726
> > > > > 		22505421080001010101010101010101
> > > > > 		01010101010128150040410026301888
> > > > > 		3600f6b800000018ed10004041002630
> > > > > 		18883600f6b9000000180000000f0061
> > > > > 		43326143280f01000daf0714000000fe
> > > > > 		004e31323158352d4c303620202000ed
> > > > > 	scaling mode: Full aspect 
> > > > > 		supported: Full, Center, Full aspect
> > > > > 	non-desktop: 0 
> > > > > 		range: (0, 1)
> > > > > 	link-status: Good 
> > > > > 		supported: Good, Bad
> > > > >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> > > > >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> > > > >   1024x768 (0x45) 65.000MHz -HSync -VSync
> > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> > > > >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> > > > >   1024x768 (0x46) 43.330MHz -HSync -VSync
> > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> > > > >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> > > > >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> > > > >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> > > > >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> > > > >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> > > > >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> > > > >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> > > > 
> > > > Where are all these dblscan modes coming from?
> > > > 
> > > > Did you add them manually or are they being automatically
> > > > generated by something?
> > > > 
> > > 
> > > This is pure automagic xserver-xorg-video-intel on i915 drm/kms.
> > > 
> > > After booting back into a working 4.16 kernel I see the same xrandr --verbose
> > > list with all the DoubleScan modes, with which xgalaga is functional.
> > > 
> > > There's this in the Xorg.log:
> > > 
> > > ```
> > > [    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
> > > [    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
> > > [    12.882] (II) intel(0): Output VGA1 has no monitor section
> > > [    12.883] (II) intel(0): EDID for output LVDS1
> > > [    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
> > > [    12.883] (II) intel(0): Year: 2006  Week: 48
> > > [    12.883] (II) intel(0): EDID Version: 1.3
> > > [    12.883] (II) intel(0): Digital Display Input
> > > [    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
> > > [    12.883] (II) intel(0): Gamma: 2.20
> > > [    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
> > > [    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
> > > [    12.883] (II) intel(0): First detailed timing is preferred mode
> > > [    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
> > > [    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
> > > [    12.883] (II) intel(0): Supported established timings:
> > > [    12.883] (II) intel(0): 640x480@60Hz
> > > [    12.883] (II) intel(0): 800x600@60Hz
> > > [    12.883] (II) intel(0): 1024x768@60Hz
> > > [    12.883] (II) intel(0): Manufacturer's mask: 0
> > > [    12.883] (II) intel(0): Supported detailed timing:
> > > [    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
> > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > [    12.883] (II) intel(0): Supported detailed timing:
> > > [    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
> > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > [    12.883] (II) intel(0): Unknown vendor-specific block f
> > > [    12.883] (II) intel(0):  N121X5-L06
> > > [    12.883] (II) intel(0): EDID (in hex):
> > > [    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
> > > [    12.883] (II) intel(0):     3010010380191278eafe609555518726
> > > [    12.883] (II) intel(0):     22505421080001010101010101010101
> > > [    12.883] (II) intel(0):     01010101010128150040410026301888
> > > [    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
> > > [    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
> > > [    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
> > > [    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
> > > [    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > [    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
> > > [    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)
> > 
> > So first it rejected them all, but somehow later they get added back?
> > Very odd.
> > 
> > Hmm. Must be coming from that default modes thing...
> > 
> > Looks pretty much like ddx bug to me. Does the following ddx patch get
> > rid of the bogus dblscan modes?
> > 
> 
> <snip>
> 
> No, that didn't seem to change anything.  After some digging, I found:
> 
> xorg-server-1.19.2/hw/xfree86/drivers/modesetting/drmmode_display.c::drmmode_output_init():
> 
> ```
> 1796     drmmode_output->output_id = mode_res->connectors[num];
> 1797     drmmode_output->mode_output = koutput;
> 1798     drmmode_output->mode_encoders = kencoders;
> 1799     drmmode_output->drmmode = drmmode;
> 1800     output->mm_width = koutput->mmWidth;
> 1801     output->mm_height = koutput->mmHeight;
> 1802 
> 1803     output->subpixel_order = subpixel_conv_table[koutput->subpixel];
> 1804     output->interlaceAllowed = TRUE;
> 1805     output->doubleScanAllowed = TRUE;
> 1806     output->driver_private = drmmode_output;
> ```

That's the modesetting ddx. Irrelevant as long as you're using the intel
ddx.

> 
> I don't see anywhere that sets interlaceAllowed or doubleScanAllowed to FALSE,
> except xserver-xorg-video-intel/src/legacy/i810/i810_driver.c.
> 
> So your change never invalidates the doublescan modes, because they're allowed
> for the output.
> 
> Assuming this can be corrected by disallowing these for the output
> somewhere appropriate, isn't this still a userspace-breaking kernel
> regression?

I suppose.

> 
> It seems to me like the kernel should just ignore the doublescan mode
> flag and continue if a non-doublescan equivalent mode exists, given the
> userspace situation this change uncovered.

The only reason why ignoring it has worked is due to the panel fitter
being used on LVDS outputs. For other output types anyone trying to
use doublescan would probabably have gotten a fail or black/bogus
output.

This is the annoying part. I guess we want to ignore it for LVDS & co.
but reject it for everything else. Which means we have to allow it
into the kernel and reject it later. I guess for now we'll want to
move the check back into the connector .mode_valid() hooks and later
on we have to figure out how we can actually use those hooks to also
validate the user mode (which has been on the TODO list for
several years).

> 
> Surely xgalaga isn't the only application using this API in such an
> obvious fashion.

Apart from xvidtune (which I don't think anyone uses anymore)
some games still use it in some fashion. Which is actually a
real pain because vidmode doesn't handle screen rotation at all. 
I kinda wish it would disappear so at least games couldn't mess
up rotated display setups so easily.

-- 
Ville Syrjälä
Intel

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 20:41                 ` Ville Syrjälä
@ 2018-05-23 21:27                   ` Vito Caputo
  2018-05-23 22:27                     ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-23 21:27 UTC (permalink / raw)
  To: Ville Syrjälä; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 11:41:25PM +0300, Ville Syrjälä wrote:
> On Wed, May 23, 2018 at 01:15:05PM -0700, Vito Caputo wrote:
> > On Wed, May 23, 2018 at 10:12:22PM +0300, Ville Syrjälä wrote:
> > > On Wed, May 23, 2018 at 11:39:00AM -0700, Vito Caputo wrote:
> > > > On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> > > > > On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > > > > > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > > > > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > > > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > > > > > Hello all,
> > > > > > > > > > 
> > > > > > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > > > > > evening.
> > > > > > > > > > 
> > > > > > > > > > > ~$ xgalaga
> > > > > > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > > > > > >   Value in failed request:  0x120004e
> > > > > > > > > > >   Serial number of failed request:  199
> > > > > > > > > > >   Current serial number in output stream:  203
> > > > > > > > > > 
> > > > > > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > > > > > 
> > > > > > > > > > Also observed, though this is surely a different regression, the game
> > > > > > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > > > > > 
> > > > > > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > > > > > 
> > > > > > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > > > > > 
> > > > > > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > > > > > playable on this machine in windowed mode when I last did.
> > > > > > > > > > 
> > > > > > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > > > > > 
> > > > > > > > > 
> > > > > > > > > Just built and booted v4.17-rc6, still broken.
> > > > > > > > 
> > > > > > > > Bisected to:
> > > > > > > > 
> > > > > > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > > > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > > > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > > > > > 
> > > > > > > >     drm/i915: Provide a device level .mode_valid() hook
> > > > > > > >     
> > > > > > > >     We never support certain mode flags etc. Reject those early on in the
> > > > > > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > > > > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > > > > > >     we never see those flags even from user mode as the
> > > > > > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > > > > > >     
> > > > > > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > > > > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > > > 
> > > > > > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > > > > > the ddx doesn't validate before passing it on to the kernel. So far
> > > > > > > I can't reproduce the problem here unfortnately.
> > > > > > > 
> > > > > > > Can you try the following patch and reproduce the problem with
> > > > > > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > > > > > modeline looks like?
> > > > > > > 
> > > > > > 
> > > > > > dmesg after xgalaga fails:
> > > > > > 
> > > > > > ```
> > > > > > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > > > > > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> > > > > 
> > > > > 0x20 == dblscan
> > > > > 
> > > > > > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > > > > > ```
> > > > > > 
> > > > > > xrandr --verbose:
> > > > > > 
> > > > > > ```
> > > > > > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > > > > > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > > > > > 	Identifier: 0x41
> > > > > > 	Timestamp:  23375
> > > > > > 	Subpixel:   horizontal rgb
> > > > > > 	Gamma:      1.0:1.0:1.0
> > > > > > 	Brightness: 1.0
> > > > > > 	Clones:    
> > > > > > 	CRTC:       0
> > > > > > 	CRTCs:      0 1
> > > > > > 	Transform:  1.000000 0.000000 0.000000
> > > > > > 	            0.000000 1.000000 0.000000
> > > > > > 	            0.000000 0.000000 1.000000
> > > > > > 	           filter: 
> > > > > > 	EDID: 
> > > > > > 		00ffffffffffff0030ae004000000000
> > > > > > 		3010010380191278eafe609555518726
> > > > > > 		22505421080001010101010101010101
> > > > > > 		01010101010128150040410026301888
> > > > > > 		3600f6b800000018ed10004041002630
> > > > > > 		18883600f6b9000000180000000f0061
> > > > > > 		43326143280f01000daf0714000000fe
> > > > > > 		004e31323158352d4c303620202000ed
> > > > > > 	scaling mode: Full aspect 
> > > > > > 		supported: Full, Center, Full aspect
> > > > > > 	non-desktop: 0 
> > > > > > 		range: (0, 1)
> > > > > > 	link-status: Good 
> > > > > > 		supported: Good, Bad
> > > > > >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> > > > > >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> > > > > >   1024x768 (0x45) 65.000MHz -HSync -VSync
> > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> > > > > >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> > > > > >   1024x768 (0x46) 43.330MHz -HSync -VSync
> > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> > > > > >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> > > > > >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> > > > > >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> > > > > >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> > > > > >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> > > > > >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> > > > > >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> > > > > 
> > > > > Where are all these dblscan modes coming from?
> > > > > 
> > > > > Did you add them manually or are they being automatically
> > > > > generated by something?
> > > > > 
> > > > 
> > > > This is pure automagic xserver-xorg-video-intel on i915 drm/kms.
> > > > 
> > > > After booting back into a working 4.16 kernel I see the same xrandr --verbose
> > > > list with all the DoubleScan modes, with which xgalaga is functional.
> > > > 
> > > > There's this in the Xorg.log:
> > > > 
> > > > ```
> > > > [    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
> > > > [    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
> > > > [    12.882] (II) intel(0): Output VGA1 has no monitor section
> > > > [    12.883] (II) intel(0): EDID for output LVDS1
> > > > [    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
> > > > [    12.883] (II) intel(0): Year: 2006  Week: 48
> > > > [    12.883] (II) intel(0): EDID Version: 1.3
> > > > [    12.883] (II) intel(0): Digital Display Input
> > > > [    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
> > > > [    12.883] (II) intel(0): Gamma: 2.20
> > > > [    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
> > > > [    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
> > > > [    12.883] (II) intel(0): First detailed timing is preferred mode
> > > > [    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
> > > > [    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
> > > > [    12.883] (II) intel(0): Supported established timings:
> > > > [    12.883] (II) intel(0): 640x480@60Hz
> > > > [    12.883] (II) intel(0): 800x600@60Hz
> > > > [    12.883] (II) intel(0): 1024x768@60Hz
> > > > [    12.883] (II) intel(0): Manufacturer's mask: 0
> > > > [    12.883] (II) intel(0): Supported detailed timing:
> > > > [    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
> > > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > > [    12.883] (II) intel(0): Supported detailed timing:
> > > > [    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
> > > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > > [    12.883] (II) intel(0): Unknown vendor-specific block f
> > > > [    12.883] (II) intel(0):  N121X5-L06
> > > > [    12.883] (II) intel(0): EDID (in hex):
> > > > [    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
> > > > [    12.883] (II) intel(0):     3010010380191278eafe609555518726
> > > > [    12.883] (II) intel(0):     22505421080001010101010101010101
> > > > [    12.883] (II) intel(0):     01010101010128150040410026301888
> > > > [    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
> > > > [    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
> > > > [    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
> > > > [    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
> > > > [    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > > [    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
> > > > [    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)
> > > 
> > > So first it rejected them all, but somehow later they get added back?
> > > Very odd.
> > > 
> > > Hmm. Must be coming from that default modes thing...
> > > 
> > > Looks pretty much like ddx bug to me. Does the following ddx patch get
> > > rid of the bogus dblscan modes?
> > > 
> > 
> > <snip>
> > 
> > No, that didn't seem to change anything.  After some digging, I found:
> > 
> > xorg-server-1.19.2/hw/xfree86/drivers/modesetting/drmmode_display.c::drmmode_output_init():
> > 
> > ```
> > 1796     drmmode_output->output_id = mode_res->connectors[num];
> > 1797     drmmode_output->mode_output = koutput;
> > 1798     drmmode_output->mode_encoders = kencoders;
> > 1799     drmmode_output->drmmode = drmmode;
> > 1800     output->mm_width = koutput->mmWidth;
> > 1801     output->mm_height = koutput->mmHeight;
> > 1802 
> > 1803     output->subpixel_order = subpixel_conv_table[koutput->subpixel];
> > 1804     output->interlaceAllowed = TRUE;
> > 1805     output->doubleScanAllowed = TRUE;
> > 1806     output->driver_private = drmmode_output;
> > ```
> 
> That's the modesetting ddx. Irrelevant as long as you're using the intel
> ddx.
> 

Hmm, that should have been obvious given the path, clearly I'm not
familiar with how this is structured.  Xorg cannot start without
modesetting_drv.so, even when using intel_drv.so, so I thought they were
more intimately sharing functionality.

I wonder why the change didn't work then; intel_output_init() doesn't
set doubleScanAllowed=TRUE and xf86OutputCreate() uses calloc() so...

If you want me to test another patch I'm happy to.  Even if the kernel
stays compatible with the bug, it still makes sense to fix userspace.

> > 
> > I don't see anywhere that sets interlaceAllowed or doubleScanAllowed to FALSE,
> > except xserver-xorg-video-intel/src/legacy/i810/i810_driver.c.
> > 
> > So your change never invalidates the doublescan modes, because they're allowed
> > for the output.
> > 
> > Assuming this can be corrected by disallowing these for the output
> > somewhere appropriate, isn't this still a userspace-breaking kernel
> > regression?
> 
> I suppose.
> 
> > 
> > It seems to me like the kernel should just ignore the doublescan mode
> > flag and continue if a non-doublescan equivalent mode exists, given the
> > userspace situation this change uncovered.
> 
> The only reason why ignoring it has worked is due to the panel fitter
> being used on LVDS outputs. For other output types anyone trying to
> use doublescan would probabably have gotten a fail or black/bogus
> output.
> 
> This is the annoying part. I guess we want to ignore it for LVDS & co.
> but reject it for everything else. Which means we have to allow it
> into the kernel and reject it later. I guess for now we'll want to
> move the check back into the connector .mode_valid() hooks and later
> on we have to figure out how we can actually use those hooks to also
> validate the user mode (which has been on the TODO list for
> several years).
> 

This sounds like a reasonable interim solution.

Thanks,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 21:27                   ` Vito Caputo
@ 2018-05-23 22:27                     ` Vito Caputo
  2018-05-24  0:21                       ` Vito Caputo
  0 siblings, 1 reply; 13+ messages in thread
From: Vito Caputo @ 2018-05-23 22:27 UTC (permalink / raw)
  To: Ville Syrjälä; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 02:27:48PM -0700, Vito Caputo wrote:
> On Wed, May 23, 2018 at 11:41:25PM +0300, Ville Syrjälä wrote:
> > On Wed, May 23, 2018 at 01:15:05PM -0700, Vito Caputo wrote:
> > > On Wed, May 23, 2018 at 10:12:22PM +0300, Ville Syrjälä wrote:
> > > > On Wed, May 23, 2018 at 11:39:00AM -0700, Vito Caputo wrote:
> > > > > On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> > > > > > On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > > > > > > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > > > > > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > > > > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > > > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > > > > > > Hello all,
> > > > > > > > > > > 
> > > > > > > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > > > > > > evening.
> > > > > > > > > > > 
> > > > > > > > > > > > ~$ xgalaga
> > > > > > > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > > > > > > >   Value in failed request:  0x120004e
> > > > > > > > > > > >   Serial number of failed request:  199
> > > > > > > > > > > >   Current serial number in output stream:  203
> > > > > > > > > > > 
> > > > > > > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > > > > > > 
> > > > > > > > > > > Also observed, though this is surely a different regression, the game
> > > > > > > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > > > > > > 
> > > > > > > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > > > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > > > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > > > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > > > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > > > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > > > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > > > > > > 
> > > > > > > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > > > > > > 
> > > > > > > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > > > > > > playable on this machine in windowed mode when I last did.
> > > > > > > > > > > 
> > > > > > > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > > > > > > 
> > > > > > > > > > 
> > > > > > > > > > Just built and booted v4.17-rc6, still broken.
> > > > > > > > > 
> > > > > > > > > Bisected to:
> > > > > > > > > 
> > > > > > > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > > > > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > > > > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > > > > > > 
> > > > > > > > >     drm/i915: Provide a device level .mode_valid() hook
> > > > > > > > >     
> > > > > > > > >     We never support certain mode flags etc. Reject those early on in the
> > > > > > > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > > > > > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > > > > > > >     we never see those flags even from user mode as the
> > > > > > > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > > > > > > >     
> > > > > > > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > > > > > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > > > > 
> > > > > > > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > > > > > > the ddx doesn't validate before passing it on to the kernel. So far
> > > > > > > > I can't reproduce the problem here unfortnately.
> > > > > > > > 
> > > > > > > > Can you try the following patch and reproduce the problem with
> > > > > > > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > > > > > > modeline looks like?
> > > > > > > > 
> > > > > > > 
> > > > > > > dmesg after xgalaga fails:
> > > > > > > 
> > > > > > > ```
> > > > > > > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > > > > > > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> > > > > > 
> > > > > > 0x20 == dblscan
> > > > > > 
> > > > > > > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > > > > > > ```
> > > > > > > 
> > > > > > > xrandr --verbose:
> > > > > > > 
> > > > > > > ```
> > > > > > > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > > > > > > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > > > > > > 	Identifier: 0x41
> > > > > > > 	Timestamp:  23375
> > > > > > > 	Subpixel:   horizontal rgb
> > > > > > > 	Gamma:      1.0:1.0:1.0
> > > > > > > 	Brightness: 1.0
> > > > > > > 	Clones:    
> > > > > > > 	CRTC:       0
> > > > > > > 	CRTCs:      0 1
> > > > > > > 	Transform:  1.000000 0.000000 0.000000
> > > > > > > 	            0.000000 1.000000 0.000000
> > > > > > > 	            0.000000 0.000000 1.000000
> > > > > > > 	           filter: 
> > > > > > > 	EDID: 
> > > > > > > 		00ffffffffffff0030ae004000000000
> > > > > > > 		3010010380191278eafe609555518726
> > > > > > > 		22505421080001010101010101010101
> > > > > > > 		01010101010128150040410026301888
> > > > > > > 		3600f6b800000018ed10004041002630
> > > > > > > 		18883600f6b9000000180000000f0061
> > > > > > > 		43326143280f01000daf0714000000fe
> > > > > > > 		004e31323158352d4c303620202000ed
> > > > > > > 	scaling mode: Full aspect 
> > > > > > > 		supported: Full, Center, Full aspect
> > > > > > > 	non-desktop: 0 
> > > > > > > 		range: (0, 1)
> > > > > > > 	link-status: Good 
> > > > > > > 		supported: Good, Bad
> > > > > > >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> > > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> > > > > > >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> > > > > > >   1024x768 (0x45) 65.000MHz -HSync -VSync
> > > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> > > > > > >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> > > > > > >   1024x768 (0x46) 43.330MHz -HSync -VSync
> > > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> > > > > > >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> > > > > > >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> > > > > > >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> > > > > > >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> > > > > > >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> > > > > > >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> > > > > > >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> > > > > > 
> > > > > > Where are all these dblscan modes coming from?
> > > > > > 
> > > > > > Did you add them manually or are they being automatically
> > > > > > generated by something?
> > > > > > 
> > > > > 
> > > > > This is pure automagic xserver-xorg-video-intel on i915 drm/kms.
> > > > > 
> > > > > After booting back into a working 4.16 kernel I see the same xrandr --verbose
> > > > > list with all the DoubleScan modes, with which xgalaga is functional.
> > > > > 
> > > > > There's this in the Xorg.log:
> > > > > 
> > > > > ```
> > > > > [    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
> > > > > [    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
> > > > > [    12.882] (II) intel(0): Output VGA1 has no monitor section
> > > > > [    12.883] (II) intel(0): EDID for output LVDS1
> > > > > [    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
> > > > > [    12.883] (II) intel(0): Year: 2006  Week: 48
> > > > > [    12.883] (II) intel(0): EDID Version: 1.3
> > > > > [    12.883] (II) intel(0): Digital Display Input
> > > > > [    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
> > > > > [    12.883] (II) intel(0): Gamma: 2.20
> > > > > [    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
> > > > > [    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
> > > > > [    12.883] (II) intel(0): First detailed timing is preferred mode
> > > > > [    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
> > > > > [    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
> > > > > [    12.883] (II) intel(0): Supported established timings:
> > > > > [    12.883] (II) intel(0): 640x480@60Hz
> > > > > [    12.883] (II) intel(0): 800x600@60Hz
> > > > > [    12.883] (II) intel(0): 1024x768@60Hz
> > > > > [    12.883] (II) intel(0): Manufacturer's mask: 0
> > > > > [    12.883] (II) intel(0): Supported detailed timing:
> > > > > [    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
> > > > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > > > [    12.883] (II) intel(0): Supported detailed timing:
> > > > > [    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
> > > > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > > > [    12.883] (II) intel(0): Unknown vendor-specific block f
> > > > > [    12.883] (II) intel(0):  N121X5-L06
> > > > > [    12.883] (II) intel(0): EDID (in hex):
> > > > > [    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
> > > > > [    12.883] (II) intel(0):     3010010380191278eafe609555518726
> > > > > [    12.883] (II) intel(0):     22505421080001010101010101010101
> > > > > [    12.883] (II) intel(0):     01010101010128150040410026301888
> > > > > [    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
> > > > > [    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
> > > > > [    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
> > > > > [    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
> > > > > [    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > > > [    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
> > > > > [    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)
> > > > 
> > > > So first it rejected them all, but somehow later they get added back?
> > > > Very odd.
> > > > 
> > > > Hmm. Must be coming from that default modes thing...
> > > > 
> > > > Looks pretty much like ddx bug to me. Does the following ddx patch get
> > > > rid of the bogus dblscan modes?
> > > > 
> > > 
> > > <snip>
> > > 
> > > No, that didn't seem to change anything.  After some digging, I found:
> > > 
> > > xorg-server-1.19.2/hw/xfree86/drivers/modesetting/drmmode_display.c::drmmode_output_init():
> > > 
> > > ```
> > > 1796     drmmode_output->output_id = mode_res->connectors[num];
> > > 1797     drmmode_output->mode_output = koutput;
> > > 1798     drmmode_output->mode_encoders = kencoders;
> > > 1799     drmmode_output->drmmode = drmmode;
> > > 1800     output->mm_width = koutput->mmWidth;
> > > 1801     output->mm_height = koutput->mmHeight;
> > > 1802 
> > > 1803     output->subpixel_order = subpixel_conv_table[koutput->subpixel];
> > > 1804     output->interlaceAllowed = TRUE;
> > > 1805     output->doubleScanAllowed = TRUE;
> > > 1806     output->driver_private = drmmode_output;
> > > ```
> > 
> > That's the modesetting ddx. Irrelevant as long as you're using the intel
> > ddx.
> > 
> 
> Hmm, that should have been obvious given the path, clearly I'm not
> familiar with how this is structured.  Xorg cannot start without
> modesetting_drv.so, even when using intel_drv.so, so I thought they were
> more intimately sharing functionality.
> 
> I wonder why the change didn't work then; intel_output_init() doesn't
> set doubleScanAllowed=TRUE and xf86OutputCreate() uses calloc() so...
> 

I just noticed something important.  I've been looking at old Xorg logs
in /var/log, it appears at some point these things moved into
~/.local/share/xorg/.

Look at the *current* logs, it looks like my xorg is using the modeset
driver.  This explains why removing modeset_drv.so breaks things.

It also explains why your xorg-video-intel patch made no difference.

As the code excerpt above illustrates, the modeset driver is enabling
both interlaceAllowed and doubleScanAllowed, so we know where it's
getting those.

Apologies for any confusion, my /var/log/ is full of Xorg.%i.log files
but they're from 2017 *sigh*.

Regards,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode
  2018-05-23 22:27                     ` Vito Caputo
@ 2018-05-24  0:21                       ` Vito Caputo
  0 siblings, 0 replies; 13+ messages in thread
From: Vito Caputo @ 2018-05-24  0:21 UTC (permalink / raw)
  To: Ville Syrjälä; +Cc: linux-kernel, daniel.vetter

On Wed, May 23, 2018 at 03:27:34PM -0700, Vito Caputo wrote:
> On Wed, May 23, 2018 at 02:27:48PM -0700, Vito Caputo wrote:
> > On Wed, May 23, 2018 at 11:41:25PM +0300, Ville Syrjälä wrote:
> > > On Wed, May 23, 2018 at 01:15:05PM -0700, Vito Caputo wrote:
> > > > On Wed, May 23, 2018 at 10:12:22PM +0300, Ville Syrjälä wrote:
> > > > > On Wed, May 23, 2018 at 11:39:00AM -0700, Vito Caputo wrote:
> > > > > > On Wed, May 23, 2018 at 09:20:37PM +0300, Ville Syrjälä wrote:
> > > > > > > On Wed, May 23, 2018 at 11:06:00AM -0700, Vito Caputo wrote:
> > > > > > > > On Wed, May 23, 2018 at 04:18:05PM +0300, Ville Syrjälä wrote:
> > > > > > > > > On Wed, May 23, 2018 at 02:49:19AM -0700, Vito Caputo wrote:
> > > > > > > > > > On Mon, May 21, 2018 at 02:57:18PM -0700, Vito Caputo wrote:
> > > > > > > > > > > On Mon, May 21, 2018 at 12:53:20PM -0700, Vito Caputo wrote:
> > > > > > > > > > > > Hello all,
> > > > > > > > > > > > 
> > > > > > > > > > > > 4.17-rc4 (my latest kernel ATM) consistently fails to start xgalaga
> > > > > > > > > > > > without -window.  I will try find time to build the latest rc this
> > > > > > > > > > > > evening.
> > > > > > > > > > > > 
> > > > > > > > > > > > > ~$ xgalaga
> > > > > > > > > > > > > X Error of failed request:  BadValue (integer parameter out of range for operation)
> > > > > > > > > > > > >   Major opcode of failed request:  152 (XFree86-VidModeExtension)
> > > > > > > > > > > > >   Minor opcode of failed request:  10 (XF86VidModeSwitchToMode)
> > > > > > > > > > > > >   Value in failed request:  0x120004e
> > > > > > > > > > > > >   Serial number of failed request:  199
> > > > > > > > > > > > >   Current serial number in output stream:  203
> > > > > > > > > > > > 
> > > > > > > > > > > > Haven't dug into this much yet, only did a perfunctory check by booting into a
> > > > > > > > > > > > few older kernels (4.11, 4.12, 4.16) and the problem is absent on all of them.
> > > > > > > > > > > > It appears to be a 4.17-specific regression right now.
> > > > > > > > > > > > 
> > > > > > > > > > > > Also observed, though this is surely a different regression, the game
> > > > > > > > > > > > ran like molasses with -window, showing some prominent kworkers in top:
> > > > > > > > > > > > 
> > > > > > > > > > > >   692 vc        20   0  312852  45884  20556 R  32.0  1.2   0:08.69 Xorg           
> > > > > > > > > > > >   102 root      20   0       0      0      0 R  11.2  0.0   0:01.43 kworker/1:3    
> > > > > > > > > > > >    94 root      20   0       0      0      0 I   8.9  0.0   0:00.83 kworker/0:2    
> > > > > > > > > > > >   696 vc        20   0   39948   4124   2912 S   1.0  0.1   0:05.57 vwm            
> > > > > > > > > > > >   902 vc        30  10   46372   4144   3500 S   0.7  0.1   0:00.08 xgalaga        
> > > > > > > > > > > >   891 vc        30  10   44924   3868   3156 R   0.3  0.1   0:00.09 top            
> > > > > > > > > > > >   903 vc        30  10    4180   1184   1100 S   0.3  0.0   0:00.01 xgal.sndsrv.oss
> > > > > > > > > > > > 
> > > > > > > > > > > > The windowed performance issue was observed on the older kernels tested
> > > > > > > > > > > > as well, though 4.11 felt better and didn't have the busy kworkers.
> > > > > > > > > > > > 
> > > > > > > > > > > > I have not attempted to play xgalaga for ages, but it used to be perfectly
> > > > > > > > > > > > playable on this machine in windowed mode when I last did.
> > > > > > > > > > > > 
> > > > > > > > > > > > Machine is the venerable Thinkpad X61s, 1.8Ghz, Debian 9, config attached.
> > > > > > > > > > > > 
> > > > > > > > > > > 
> > > > > > > > > > > Just built and booted v4.17-rc6, still broken.
> > > > > > > > > > 
> > > > > > > > > > Bisected to:
> > > > > > > > > > 
> > > > > > > > > > e995ca0b8139c5f6807095464e969931b443f55a is the first bad commit
> > > > > > > > > > commit e995ca0b8139c5f6807095464e969931b443f55a
> > > > > > > > > > Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > > > > Date:   Tue Nov 14 20:32:58 2017 +0200
> > > > > > > > > > 
> > > > > > > > > >     drm/i915: Provide a device level .mode_valid() hook
> > > > > > > > > >     
> > > > > > > > > >     We never support certain mode flags etc. Reject those early on in the
> > > > > > > > > >     mode_config.mode_valid() hook. That allows us to remove some duplicated
> > > > > > > > > >     checks from the connector .mode_valid() hooks, and it guarantees that
> > > > > > > > > >     we never see those flags even from user mode as the
> > > > > > > > > >     mode_config.mode_valid() hooks gets executed for those as well.
> > > > > > > > > >     
> > > > > > > > > >     Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > > > > >     Link: https://patchwork.freedesktop.org/patch/msgid/20171114183258.16976-11-ville.syrjala@linux.intel.com
> > > > > > > > > >     Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > > > > > > > 
> > > > > > > > > Hmm. I guess xgalaga passes some garbage in via xf86vidmode which
> > > > > > > > > the ddx doesn't validate before passing it on to the kernel. So far
> > > > > > > > > I can't reproduce the problem here unfortnately.
> > > > > > > > > 
> > > > > > > > > Can you try the following patch and reproduce the problem with
> > > > > > > > > drm.debug=0xe passed to the kernel so that we can seewhat the bad
> > > > > > > > > modeline looks like?
> > > > > > > > > 
> > > > > > > > 
> > > > > > > > dmesg after xgalaga fails:
> > > > > > > > 
> > > > > > > > ```
> > > > > > > > [   75.617448] [drm:drm_mode_convert_umode] Bad user mode:
> > > > > > > > [   75.617455] [drm:drm_mode_debug_printmodeline] Modeline 57:"800x600" 0 81000 800 832 928 1080 600 600 602 625 0x0 0x25
> > > > > > > 
> > > > > > > 0x20 == dblscan
> > > > > > > 
> > > > > > > > [   75.617458] [drm:drm_mode_setcrtc] Invalid mode
> > > > > > > > ```
> > > > > > > > 
> > > > > > > > xrandr --verbose:
> > > > > > > > 
> > > > > > > > ```
> > > > > > > > Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
> > > > > > > > LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
> > > > > > > > 	Identifier: 0x41
> > > > > > > > 	Timestamp:  23375
> > > > > > > > 	Subpixel:   horizontal rgb
> > > > > > > > 	Gamma:      1.0:1.0:1.0
> > > > > > > > 	Brightness: 1.0
> > > > > > > > 	Clones:    
> > > > > > > > 	CRTC:       0
> > > > > > > > 	CRTCs:      0 1
> > > > > > > > 	Transform:  1.000000 0.000000 0.000000
> > > > > > > > 	            0.000000 1.000000 0.000000
> > > > > > > > 	            0.000000 0.000000 1.000000
> > > > > > > > 	           filter: 
> > > > > > > > 	EDID: 
> > > > > > > > 		00ffffffffffff0030ae004000000000
> > > > > > > > 		3010010380191278eafe609555518726
> > > > > > > > 		22505421080001010101010101010101
> > > > > > > > 		01010101010128150040410026301888
> > > > > > > > 		3600f6b800000018ed10004041002630
> > > > > > > > 		18883600f6b9000000180000000f0061
> > > > > > > > 		43326143280f01000daf0714000000fe
> > > > > > > > 		004e31323158352d4c303620202000ed
> > > > > > > > 	scaling mode: Full aspect 
> > > > > > > > 		supported: Full, Center, Full aspect
> > > > > > > > 	non-desktop: 0 
> > > > > > > > 		range: (0, 1)
> > > > > > > > 	link-status: Good 
> > > > > > > > 		supported: Good, Bad
> > > > > > > >   1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
> > > > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
> > > > > > > >         v: height  768 start  771 end  777 total  806           clock  50.00Hz
> > > > > > > >   1024x768 (0x45) 65.000MHz -HSync -VSync
> > > > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
> > > > > > > >         v: height  768 start  771 end  777 total  806           clock  60.00Hz
> > > > > > > >   1024x768 (0x46) 43.330MHz -HSync -VSync
> > > > > > > >         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
> > > > > > > >         v: height  768 start  771 end  777 total  806           clock  40.00Hz
> > > > > > > >   960x720 (0x47) 117.000MHz -HSync +VSync DoubleScan
> > > > > > > >         h: width   960 start 1024 end 1128 total 1300 skew    0 clock  90.00KHz
> > > > > > > >         v: height  720 start  720 end  722 total  750           clock  60.00Hz
> > > > > > > >   928x696 (0x48) 109.150MHz -HSync +VSync DoubleScan
> > > > > > > >         h: width   928 start  976 end 1088 total 1264 skew    0 clock  86.35KHz
> > > > > > > >         v: height  696 start  696 end  698 total  719           clock  60.05Hz
> > > > > > > 
> > > > > > > Where are all these dblscan modes coming from?
> > > > > > > 
> > > > > > > Did you add them manually or are they being automatically
> > > > > > > generated by something?
> > > > > > > 
> > > > > > 
> > > > > > This is pure automagic xserver-xorg-video-intel on i915 drm/kms.
> > > > > > 
> > > > > > After booting back into a working 4.16 kernel I see the same xrandr --verbose
> > > > > > list with all the DoubleScan modes, with which xgalaga is functional.
> > > > > > 
> > > > > > There's this in the Xorg.log:
> > > > > > 
> > > > > > ```
> > > > > > [    12.856] (II) intel(0): Output LVDS1 using monitor section Monitor0
> > > > > > [    12.857] (--) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
> > > > > > [    12.882] (II) intel(0): Output VGA1 has no monitor section
> > > > > > [    12.883] (II) intel(0): EDID for output LVDS1
> > > > > > [    12.883] (II) intel(0): Manufacturer: LEN  Model: 4000  Serial#: 0
> > > > > > [    12.883] (II) intel(0): Year: 2006  Week: 48
> > > > > > [    12.883] (II) intel(0): EDID Version: 1.3
> > > > > > [    12.883] (II) intel(0): Digital Display Input
> > > > > > [    12.883] (II) intel(0): Max Image Size [cm]: horiz.: 25  vert.: 18
> > > > > > [    12.883] (II) intel(0): Gamma: 2.20
> > > > > > [    12.883] (II) intel(0): DPMS capabilities: StandBy Suspend Off
> > > > > > [    12.883] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4 
> > > > > > [    12.883] (II) intel(0): First detailed timing is preferred mode
> > > > > > [    12.883] (II) intel(0): redX: 0.585 redY: 0.335   greenX: 0.319 greenY: 0.529
> > > > > > [    12.883] (II) intel(0): blueX: 0.149 blueY: 0.135   whiteX: 0.312 whiteY: 0.328
> > > > > > [    12.883] (II) intel(0): Supported established timings:
> > > > > > [    12.883] (II) intel(0): 640x480@60Hz
> > > > > > [    12.883] (II) intel(0): 800x600@60Hz
> > > > > > [    12.883] (II) intel(0): 1024x768@60Hz
> > > > > > [    12.883] (II) intel(0): Manufacturer's mask: 0
> > > > > > [    12.883] (II) intel(0): Supported detailed timing:
> > > > > > [    12.883] (II) intel(0): clock: 54.2 MHz   Image Size:  246 x 184 mm
> > > > > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > > > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > > > > [    12.883] (II) intel(0): Supported detailed timing:
> > > > > > [    12.883] (II) intel(0): clock: 43.3 MHz   Image Size:  246 x 185 mm
> > > > > > [    12.883] (II) intel(0): h_active: 1024  h_sync: 1048  h_sync_end 1184 h_blank_end 1344 h_border: 0
> > > > > > [    12.883] (II) intel(0): v_active: 768  v_sync: 771  v_sync_end 777 v_blanking: 806 v_border: 0
> > > > > > [    12.883] (II) intel(0): Unknown vendor-specific block f
> > > > > > [    12.883] (II) intel(0):  N121X5-L06
> > > > > > [    12.883] (II) intel(0): EDID (in hex):
> > > > > > [    12.883] (II) intel(0):     00ffffffffffff0030ae004000000000
> > > > > > [    12.883] (II) intel(0):     3010010380191278eafe609555518726
> > > > > > [    12.883] (II) intel(0):     22505421080001010101010101010101
> > > > > > [    12.883] (II) intel(0):     01010101010128150040410026301888
> > > > > > [    12.883] (II) intel(0):     3600f6b800000018ed10004041002630
> > > > > > [    12.883] (II) intel(0):     18883600f6b9000000180000000f0061
> > > > > > [    12.883] (II) intel(0):     43326143280f01000daf0714000000fe
> > > > > > [    12.883] (II) intel(0):     004e31323158352d4c303620202000ed
> > > > > > [    12.883] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "576x432" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "680x384" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "720x450" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "800x512" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > > > > [    12.883] (II) intel(0): Not using default mode "840x525" (doublescan mode not supported)
> > > > > > [    12.884] (II) intel(0): Not using default mode "960x540" (doublescan mode not supported)
> > > > > > [    12.884] (II) intel(0): Not using default mode "960x600" (doublescan mode not supported)
> > > > > 
> > > > > So first it rejected them all, but somehow later they get added back?
> > > > > Very odd.
> > > > > 
> > > > > Hmm. Must be coming from that default modes thing...
> > > > > 
> > > > > Looks pretty much like ddx bug to me. Does the following ddx patch get
> > > > > rid of the bogus dblscan modes?
> > > > > 
> > > > 
> > > > <snip>
> > > > 
> > > > No, that didn't seem to change anything.  After some digging, I found:
> > > > 
> > > > xorg-server-1.19.2/hw/xfree86/drivers/modesetting/drmmode_display.c::drmmode_output_init():
> > > > 
> > > > ```
> > > > 1796     drmmode_output->output_id = mode_res->connectors[num];
> > > > 1797     drmmode_output->mode_output = koutput;
> > > > 1798     drmmode_output->mode_encoders = kencoders;
> > > > 1799     drmmode_output->drmmode = drmmode;
> > > > 1800     output->mm_width = koutput->mmWidth;
> > > > 1801     output->mm_height = koutput->mmHeight;
> > > > 1802 
> > > > 1803     output->subpixel_order = subpixel_conv_table[koutput->subpixel];
> > > > 1804     output->interlaceAllowed = TRUE;
> > > > 1805     output->doubleScanAllowed = TRUE;
> > > > 1806     output->driver_private = drmmode_output;
> > > > ```
> > > 
> > > That's the modesetting ddx. Irrelevant as long as you're using the intel
> > > ddx.
> > > 
> > 
> > Hmm, that should have been obvious given the path, clearly I'm not
> > familiar with how this is structured.  Xorg cannot start without
> > modesetting_drv.so, even when using intel_drv.so, so I thought they were
> > more intimately sharing functionality.
> > 
> > I wonder why the change didn't work then; intel_output_init() doesn't
> > set doubleScanAllowed=TRUE and xf86OutputCreate() uses calloc() so...
> > 
> 
> I just noticed something important.  I've been looking at old Xorg logs
> in /var/log, it appears at some point these things moved into
> ~/.local/share/xorg/.
> 
> Look at the *current* logs, it looks like my xorg is using the modeset
> driver.  This explains why removing modeset_drv.so breaks things.
> 
> It also explains why your xorg-video-intel patch made no difference.
> 
> As the code excerpt above illustrates, the modeset driver is enabling
> both interlaceAllowed and doubleScanAllowed, so we know where it's
> getting those.
> 
> Apologies for any confusion, my /var/log/ is full of Xorg.%i.log files
> but they're from 2017 *sigh*.
> 

I've just built and tested the modesetting driver with:
-	output->doubleScanAllowed = TRUE;
+	output->doubleScanAllowed = FALSE;

and it fixes xgalaga on 4.17-rc6.

xrandr --verbose output is now:

```
Screen 0: minimum 320 x 200, current 1024 x 768, maximum 8192 x 8192
LVDS-1 connected primary 1024x768+0+0 (0x44) normal (normal left inverted right x axis y axis) 246mm x 184mm
        Identifier: 0x41
        Timestamp:  16465276
        Subpixel:   horizontal rgb
        Gamma:      1.0:1.0:1.0
        Brightness: 1.0
        Clones:    
        CRTC:       0
        CRTCs:      0 1
        Transform:  1.000000 0.000000 0.000000
                    0.000000 1.000000 0.000000
                    0.000000 0.000000 1.000000
                   filter: 
        EDID: 
                00ffffffffffff0030ae004000000000
                3010010380191278eafe609555518726
                22505421080001010101010101010101
                01010101010128150040410026301888
                3600f6b800000018ed10004041002630
                18883600f6b9000000180000000f0061
                43326143280f01000daf0714000000fe
                004e31323158352d4c303620202000ed
        scaling mode: Full aspect 
                supported: Full, Center, Full aspect
        non-desktop: 0 
                range: (0, 1)
        link-status: Good 
                supported: Good, Bad
  1024x768 (0x44) 54.160MHz -HSync -VSync *current +preferred
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  40.30KHz
        v: height  768 start  771 end  777 total  806           clock  50.00Hz
  1024x768 (0x45) 65.000MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  48.36KHz
        v: height  768 start  771 end  777 total  806           clock  60.00Hz
  1024x768 (0x46) 43.330MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock  32.24KHz
        v: height  768 start  771 end  777 total  806           clock  40.00Hz
  800x600 (0x47) 40.000MHz +HSync +VSync
        h: width   800 start  840 end  968 total 1056 skew    0 clock  37.88KHz
        v: height  600 start  601 end  605 total  628           clock  60.32Hz
  800x600 (0x48) 36.000MHz +HSync +VSync
        h: width   800 start  824 end  896 total 1024 skew    0 clock  35.16KHz
        v: height  600 start  601 end  603 total  625           clock  56.25Hz
  640x480 (0x49) 25.175MHz -HSync -VSync
        h: width   640 start  656 end  752 total  800 skew    0 clock  31.47KHz
        v: height  480 start  490 end  492 total  525           clock  59.94Hz
VGA-1 disconnected (normal left inverted right x axis y axis)
        Identifier: 0x42
        Timestamp:  16465276
        Subpixel:   unknown
        Clones:    
        CRTCs:      0 1
        Transform:  1.000000 0.000000 0.000000
                    0.000000 1.000000 0.000000
                    0.000000 0.000000 1.000000
                   filter: 
        non-desktop: 0 
                range: (0, 1)
        link-status: Good 
                supported: Good, Bad
```

So if you're going to fix the xserver-xorg-video-intel driver, it looks like
the modesetting driver should also be corrected in the same way.

Now to figure out why Debian doesn't package modesetting_drv.so as 
xserver-xorg-video-$variant, which results in requiring something like
xserver-xorg-video-intel even when modesetting is in use.

Regards,
Vito Caputo

^ permalink raw reply	[flat|nested] 13+ messages in thread

end of thread, other threads:[~2018-05-24  0:21 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-05-21 19:53 [REGRESSION] v4.17-rc4: xgalaga fails to start in fullscreen (default) mode Vito Caputo
2018-05-21 21:57 ` Vito Caputo
2018-05-23  9:49   ` Vito Caputo
2018-05-23 13:18     ` Ville Syrjälä
2018-05-23 18:06       ` Vito Caputo
2018-05-23 18:20         ` Ville Syrjälä
2018-05-23 18:39           ` Vito Caputo
2018-05-23 19:12             ` Ville Syrjälä
2018-05-23 20:15               ` Vito Caputo
2018-05-23 20:41                 ` Ville Syrjälä
2018-05-23 21:27                   ` Vito Caputo
2018-05-23 22:27                     ` Vito Caputo
2018-05-24  0:21                       ` Vito Caputo

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox