diff mbox series

[4/4] drm/i915: Rewrite CL/CTG L-shaped memory detection

Message ID 20210421153401.13847-5-ville.syrjala@linux.intel.com (mailing list archive)
State New, archived
Headers show
Series drm/i915: Fix older platforms | expand

Commit Message

Ville Syrjälä April 21, 2021, 3:34 p.m. UTC
From: Ville Syrjälä <ville.syrjala@linux.intel.com>

Currently we try to detect a symmetric memory configurations
using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
either only set on a very specific subset of machines or it
just does not exist (it's not mentioned in any public chipset
datasheets I've found). As it happens my CL/CTG machines never
set said bit, even if I populate the channels with identical
sticks.

So let's do the L-shaped memory detection the same way as the
desktop variants, ie. just look at the DRAM rank boundary
registers to see if both channels have an identical size.

With this my CL/CTG no longer claim L-shaped memory when I use
identical sticks. Also tested with non-matching sticks just to
make sure the L-shaped memory is still properly detected.

And for completeness let's update the debugfs code to dump
the correct set of registers on each platform.

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c | 15 ++++++++-------
 drivers/gpu/drm/i915/i915_debugfs.c          | 16 ++++++++++++----
 drivers/gpu/drm/i915/i915_reg.h              |  4 ++++
 3 files changed, 24 insertions(+), 11 deletions(-)

Comments

Daniel Vetter April 22, 2021, 9:49 a.m. UTC | #1
On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> 
> Currently we try to detect a symmetric memory configurations
> using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> either only set on a very specific subset of machines or it
> just does not exist (it's not mentioned in any public chipset
> datasheets I've found). As it happens my CL/CTG machines never
> set said bit, even if I populate the channels with identical
> sticks.
> 
> So let's do the L-shaped memory detection the same way as the
> desktop variants, ie. just look at the DRAM rank boundary
> registers to see if both channels have an identical size.
> 
> With this my CL/CTG no longer claim L-shaped memory when I use
> identical sticks. Also tested with non-matching sticks just to
> make sure the L-shaped memory is still properly detected.
> 
> And for completeness let's update the debugfs code to dump
> the correct set of registers on each platform.
> 
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>

Did you check this with the swapping igt? I have some vague memories of
bug reports where somehow the machine was acting like it's L-shaped memory
despite that banks were populated equally. I've iirc tried all kinds of
tricks to figure it out, all to absolutely no avail.

tbh I'd just not touch this, not really worth it.
-Daniel
> ---
>  drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c | 15 ++++++++-------
>  drivers/gpu/drm/i915/i915_debugfs.c          | 16 ++++++++++++----
>  drivers/gpu/drm/i915/i915_reg.h              |  4 ++++
>  3 files changed, 24 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> index 0fa6c38893f7..754f20768de5 100644
> --- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> +++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> @@ -693,14 +693,15 @@ static void detect_bit_6_swizzle(struct i915_ggtt *ggtt)
>  				swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
>  				swizzle_y = I915_BIT_6_SWIZZLE_9_17;
>  			}
> -			break;
> -		}
>  
> -		/* check for L-shaped memory aka modified enhanced addressing */
> -		if (IS_GEN(i915, 4) &&
> -		    !(intel_uncore_read(uncore, DCC2) & DCC2_MODIFIED_ENHANCED_DISABLE)) {
> -			swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
> -			swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
> +			/* check for L-shaped memory aka modified enhanced addressing */
> +			if (IS_GEN(i915, 4) &&
> +			    intel_uncore_read16(uncore, C0DRB3_CL) !=
> +			    intel_uncore_read16(uncore, C1DRB3_CL)) {
> +				swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
> +				swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
> +			}
> +			break;
>  		}
>  
>  		if (dcc == 0xffffffff) {
> diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
> index 8dd374691102..6de11ffcde38 100644
> --- a/drivers/gpu/drm/i915/i915_debugfs.c
> +++ b/drivers/gpu/drm/i915/i915_debugfs.c
> @@ -621,10 +621,18 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
>  			   intel_uncore_read(uncore, DCC));
>  		seq_printf(m, "DDC2 = 0x%08x\n",
>  			   intel_uncore_read(uncore, DCC2));
> -		seq_printf(m, "C0DRB3 = 0x%04x\n",
> -			   intel_uncore_read16(uncore, C0DRB3_BW));
> -		seq_printf(m, "C1DRB3 = 0x%04x\n",
> -			   intel_uncore_read16(uncore, C1DRB3_BW));
> +
> +		if (IS_G45(dev_priv) || IS_I965G(dev_priv) || IS_G33(dev_priv)) {
> +			seq_printf(m, "C0DRB3 = 0x%04x\n",
> +				   intel_uncore_read16(uncore, C0DRB3_BW));
> +			seq_printf(m, "C1DRB3 = 0x%04x\n",
> +				   intel_uncore_read16(uncore, C1DRB3_BW));
> +		} else if (IS_GEN(dev_priv, 4)) {
> +			seq_printf(m, "C0DRB3 = 0x%04x\n",
> +				   intel_uncore_read16(uncore, C0DRB3_CL));
> +			seq_printf(m, "C1DRB3 = 0x%04x\n",
> +				   intel_uncore_read16(uncore, C1DRB3_CL));
> +		}
>  	} else if (INTEL_GEN(dev_priv) >= 6) {
>  		seq_printf(m, "MAD_DIMM_C0 = 0x%08x\n",
>  			   intel_uncore_read(uncore, MAD_DIMM_C0));
> diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
> index 0587b2455ea1..055c258179a1 100644
> --- a/drivers/gpu/drm/i915/i915_reg.h
> +++ b/drivers/gpu/drm/i915/i915_reg.h
> @@ -3790,6 +3790,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
>  #define C0DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x206)
>  #define C1DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x606)
>  
> +/* 965gm,ctg DRAM channel configuration */
> +#define C0DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1206)
> +#define C1DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1306)
> +
>  /* snb MCH registers for reading the DRAM channel configuration */
>  #define MAD_DIMM_C0			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5004)
>  #define MAD_DIMM_C1			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5008)
> -- 
> 2.26.3
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Ville Syrjälä April 22, 2021, 1:11 p.m. UTC | #2
On Thu, Apr 22, 2021 at 11:49:43AM +0200, Daniel Vetter wrote:
> On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > 
> > Currently we try to detect a symmetric memory configurations
> > using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> > either only set on a very specific subset of machines or it
> > just does not exist (it's not mentioned in any public chipset
> > datasheets I've found). As it happens my CL/CTG machines never
> > set said bit, even if I populate the channels with identical
> > sticks.
> > 
> > So let's do the L-shaped memory detection the same way as the
> > desktop variants, ie. just look at the DRAM rank boundary
> > registers to see if both channels have an identical size.
> > 
> > With this my CL/CTG no longer claim L-shaped memory when I use
> > identical sticks. Also tested with non-matching sticks just to
> > make sure the L-shaped memory is still properly detected.
> > 
> > And for completeness let's update the debugfs code to dump
> > the correct set of registers on each platform.
> > 
> > Cc: Chris Wilson <chris@chris-wilson.co.uk>
> > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> 
> Did you check this with the swapping igt? I have some vague memories of
> bug reports where somehow the machine was acting like it's L-shaped memory
> despite that banks were populated equally. I've iirc tried all kinds of
> tricks to figure it out, all to absolutely no avail.

Did you have a specific test in mind? I ran a bunch of things
that seemed swizzle related. All passed just fine.

Chris did have similar concerns and suggested we should have
better tests. I guess what I should try to do is some selftests
which make sure we test both high and low physical addresses
and check the swizzle pattern is as expected. But haven't 
found the time to do that yet.

> 
> tbh I'd just not touch this, not really worth it.

It's totally worth it to get gen4 machines working again.


> -Daniel
> > ---
> >  drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c | 15 ++++++++-------
> >  drivers/gpu/drm/i915/i915_debugfs.c          | 16 ++++++++++++----
> >  drivers/gpu/drm/i915/i915_reg.h              |  4 ++++
> >  3 files changed, 24 insertions(+), 11 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> > index 0fa6c38893f7..754f20768de5 100644
> > --- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> > +++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> > @@ -693,14 +693,15 @@ static void detect_bit_6_swizzle(struct i915_ggtt *ggtt)
> >  				swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
> >  				swizzle_y = I915_BIT_6_SWIZZLE_9_17;
> >  			}
> > -			break;
> > -		}
> >  
> > -		/* check for L-shaped memory aka modified enhanced addressing */
> > -		if (IS_GEN(i915, 4) &&
> > -		    !(intel_uncore_read(uncore, DCC2) & DCC2_MODIFIED_ENHANCED_DISABLE)) {
> > -			swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
> > -			swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
> > +			/* check for L-shaped memory aka modified enhanced addressing */
> > +			if (IS_GEN(i915, 4) &&
> > +			    intel_uncore_read16(uncore, C0DRB3_CL) !=
> > +			    intel_uncore_read16(uncore, C1DRB3_CL)) {
> > +				swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
> > +				swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
> > +			}
> > +			break;
> >  		}
> >  
> >  		if (dcc == 0xffffffff) {
> > diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
> > index 8dd374691102..6de11ffcde38 100644
> > --- a/drivers/gpu/drm/i915/i915_debugfs.c
> > +++ b/drivers/gpu/drm/i915/i915_debugfs.c
> > @@ -621,10 +621,18 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
> >  			   intel_uncore_read(uncore, DCC));
> >  		seq_printf(m, "DDC2 = 0x%08x\n",
> >  			   intel_uncore_read(uncore, DCC2));
> > -		seq_printf(m, "C0DRB3 = 0x%04x\n",
> > -			   intel_uncore_read16(uncore, C0DRB3_BW));
> > -		seq_printf(m, "C1DRB3 = 0x%04x\n",
> > -			   intel_uncore_read16(uncore, C1DRB3_BW));
> > +
> > +		if (IS_G45(dev_priv) || IS_I965G(dev_priv) || IS_G33(dev_priv)) {
> > +			seq_printf(m, "C0DRB3 = 0x%04x\n",
> > +				   intel_uncore_read16(uncore, C0DRB3_BW));
> > +			seq_printf(m, "C1DRB3 = 0x%04x\n",
> > +				   intel_uncore_read16(uncore, C1DRB3_BW));
> > +		} else if (IS_GEN(dev_priv, 4)) {
> > +			seq_printf(m, "C0DRB3 = 0x%04x\n",
> > +				   intel_uncore_read16(uncore, C0DRB3_CL));
> > +			seq_printf(m, "C1DRB3 = 0x%04x\n",
> > +				   intel_uncore_read16(uncore, C1DRB3_CL));
> > +		}
> >  	} else if (INTEL_GEN(dev_priv) >= 6) {
> >  		seq_printf(m, "MAD_DIMM_C0 = 0x%08x\n",
> >  			   intel_uncore_read(uncore, MAD_DIMM_C0));
> > diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
> > index 0587b2455ea1..055c258179a1 100644
> > --- a/drivers/gpu/drm/i915/i915_reg.h
> > +++ b/drivers/gpu/drm/i915/i915_reg.h
> > @@ -3790,6 +3790,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
> >  #define C0DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x206)
> >  #define C1DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x606)
> >  
> > +/* 965gm,ctg DRAM channel configuration */
> > +#define C0DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1206)
> > +#define C1DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1306)
> > +
> >  /* snb MCH registers for reading the DRAM channel configuration */
> >  #define MAD_DIMM_C0			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5004)
> >  #define MAD_DIMM_C1			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5008)
> > -- 
> > 2.26.3
> > 
> > _______________________________________________
> > Intel-gfx mailing list
> > Intel-gfx@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> 
> -- 
> Daniel Vetter
> Software Engineer, Intel Corporation
> http://blog.ffwll.ch
Ville Syrjälä April 22, 2021, 6:51 p.m. UTC | #3
On Thu, Apr 22, 2021 at 11:49:43AM +0200, Daniel Vetter wrote:
> On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > 
> > Currently we try to detect a symmetric memory configurations
> > using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> > either only set on a very specific subset of machines or it
> > just does not exist (it's not mentioned in any public chipset
> > datasheets I've found). As it happens my CL/CTG machines never
> > set said bit, even if I populate the channels with identical
> > sticks.
> > 
> > So let's do the L-shaped memory detection the same way as the
> > desktop variants, ie. just look at the DRAM rank boundary
> > registers to see if both channels have an identical size.
> > 
> > With this my CL/CTG no longer claim L-shaped memory when I use
> > identical sticks. Also tested with non-matching sticks just to
> > make sure the L-shaped memory is still properly detected.
> > 
> > And for completeness let's update the debugfs code to dump
> > the correct set of registers on each platform.
> > 
> > Cc: Chris Wilson <chris@chris-wilson.co.uk>
> > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> 
> Did you check this with the swapping igt? I have some vague memories of
> bug reports where somehow the machine was acting like it's L-shaped memory
> despite that banks were populated equally. I've iirc tried all kinds of
> tricks to figure it out, all to absolutely no avail.

BTW looking at the patches/dumps in eg.
https://bugs.freedesktop.org/show_bug.cgi?id=28813
I can't immediately see a single thing that is actually using
the correct register offsets for cl/ctg. So I'm a bit sceptical
about how well this was researched in the past.
Daniel Vetter April 26, 2021, 4:08 p.m. UTC | #4
On Thu, Apr 22, 2021 at 04:11:22PM +0300, Ville Syrjälä wrote:
> On Thu, Apr 22, 2021 at 11:49:43AM +0200, Daniel Vetter wrote:
> > On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> > > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > 
> > > Currently we try to detect a symmetric memory configurations
> > > using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> > > either only set on a very specific subset of machines or it
> > > just does not exist (it's not mentioned in any public chipset
> > > datasheets I've found). As it happens my CL/CTG machines never
> > > set said bit, even if I populate the channels with identical
> > > sticks.
> > > 
> > > So let's do the L-shaped memory detection the same way as the
> > > desktop variants, ie. just look at the DRAM rank boundary
> > > registers to see if both channels have an identical size.
> > > 
> > > With this my CL/CTG no longer claim L-shaped memory when I use
> > > identical sticks. Also tested with non-matching sticks just to
> > > make sure the L-shaped memory is still properly detected.
> > > 
> > > And for completeness let's update the debugfs code to dump
> > > the correct set of registers on each platform.
> > > 
> > > Cc: Chris Wilson <chris@chris-wilson.co.uk>
> > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > 
> > Did you check this with the swapping igt? I have some vague memories of
> > bug reports where somehow the machine was acting like it's L-shaped memory
> > despite that banks were populated equally. I've iirc tried all kinds of
> > tricks to figure it out, all to absolutely no avail.
> 
> Did you have a specific test in mind? I ran a bunch of things
> that seemed swizzle related. All passed just fine.

gem_tiled_swapping should be the one. It tries to cycle your entire system
memory through tiled buffers into swap and out of it.
-Daniel

> 
> Chris did have similar concerns and suggested we should have
> better tests. I guess what I should try to do is some selftests
> which make sure we test both high and low physical addresses
> and check the swizzle pattern is as expected. But haven't 
> found the time to do that yet.
> 
> > 
> > tbh I'd just not touch this, not really worth it.
> 
> It's totally worth it to get gen4 machines working again.
> 
> 
> > -Daniel
> > > ---
> > >  drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c | 15 ++++++++-------
> > >  drivers/gpu/drm/i915/i915_debugfs.c          | 16 ++++++++++++----
> > >  drivers/gpu/drm/i915/i915_reg.h              |  4 ++++
> > >  3 files changed, 24 insertions(+), 11 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> > > index 0fa6c38893f7..754f20768de5 100644
> > > --- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> > > +++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
> > > @@ -693,14 +693,15 @@ static void detect_bit_6_swizzle(struct i915_ggtt *ggtt)
> > >  				swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
> > >  				swizzle_y = I915_BIT_6_SWIZZLE_9_17;
> > >  			}
> > > -			break;
> > > -		}
> > >  
> > > -		/* check for L-shaped memory aka modified enhanced addressing */
> > > -		if (IS_GEN(i915, 4) &&
> > > -		    !(intel_uncore_read(uncore, DCC2) & DCC2_MODIFIED_ENHANCED_DISABLE)) {
> > > -			swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
> > > -			swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
> > > +			/* check for L-shaped memory aka modified enhanced addressing */
> > > +			if (IS_GEN(i915, 4) &&
> > > +			    intel_uncore_read16(uncore, C0DRB3_CL) !=
> > > +			    intel_uncore_read16(uncore, C1DRB3_CL)) {
> > > +				swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
> > > +				swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
> > > +			}
> > > +			break;
> > >  		}
> > >  
> > >  		if (dcc == 0xffffffff) {
> > > diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
> > > index 8dd374691102..6de11ffcde38 100644
> > > --- a/drivers/gpu/drm/i915/i915_debugfs.c
> > > +++ b/drivers/gpu/drm/i915/i915_debugfs.c
> > > @@ -621,10 +621,18 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
> > >  			   intel_uncore_read(uncore, DCC));
> > >  		seq_printf(m, "DDC2 = 0x%08x\n",
> > >  			   intel_uncore_read(uncore, DCC2));
> > > -		seq_printf(m, "C0DRB3 = 0x%04x\n",
> > > -			   intel_uncore_read16(uncore, C0DRB3_BW));
> > > -		seq_printf(m, "C1DRB3 = 0x%04x\n",
> > > -			   intel_uncore_read16(uncore, C1DRB3_BW));
> > > +
> > > +		if (IS_G45(dev_priv) || IS_I965G(dev_priv) || IS_G33(dev_priv)) {
> > > +			seq_printf(m, "C0DRB3 = 0x%04x\n",
> > > +				   intel_uncore_read16(uncore, C0DRB3_BW));
> > > +			seq_printf(m, "C1DRB3 = 0x%04x\n",
> > > +				   intel_uncore_read16(uncore, C1DRB3_BW));
> > > +		} else if (IS_GEN(dev_priv, 4)) {
> > > +			seq_printf(m, "C0DRB3 = 0x%04x\n",
> > > +				   intel_uncore_read16(uncore, C0DRB3_CL));
> > > +			seq_printf(m, "C1DRB3 = 0x%04x\n",
> > > +				   intel_uncore_read16(uncore, C1DRB3_CL));
> > > +		}
> > >  	} else if (INTEL_GEN(dev_priv) >= 6) {
> > >  		seq_printf(m, "MAD_DIMM_C0 = 0x%08x\n",
> > >  			   intel_uncore_read(uncore, MAD_DIMM_C0));
> > > diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
> > > index 0587b2455ea1..055c258179a1 100644
> > > --- a/drivers/gpu/drm/i915/i915_reg.h
> > > +++ b/drivers/gpu/drm/i915/i915_reg.h
> > > @@ -3790,6 +3790,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
> > >  #define C0DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x206)
> > >  #define C1DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x606)
> > >  
> > > +/* 965gm,ctg DRAM channel configuration */
> > > +#define C0DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1206)
> > > +#define C1DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1306)
> > > +
> > >  /* snb MCH registers for reading the DRAM channel configuration */
> > >  #define MAD_DIMM_C0			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5004)
> > >  #define MAD_DIMM_C1			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5008)
> > > -- 
> > > 2.26.3
> > > 
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx@lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > 
> > -- 
> > Daniel Vetter
> > Software Engineer, Intel Corporation
> > http://blog.ffwll.ch
> 
> -- 
> Ville Syrjälä
> Intel
Ville Syrjälä April 26, 2021, 5:18 p.m. UTC | #5
On Mon, Apr 26, 2021 at 06:08:59PM +0200, Daniel Vetter wrote:
> On Thu, Apr 22, 2021 at 04:11:22PM +0300, Ville Syrjälä wrote:
> > On Thu, Apr 22, 2021 at 11:49:43AM +0200, Daniel Vetter wrote:
> > > On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> > > > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > 
> > > > Currently we try to detect a symmetric memory configurations
> > > > using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> > > > either only set on a very specific subset of machines or it
> > > > just does not exist (it's not mentioned in any public chipset
> > > > datasheets I've found). As it happens my CL/CTG machines never
> > > > set said bit, even if I populate the channels with identical
> > > > sticks.
> > > > 
> > > > So let's do the L-shaped memory detection the same way as the
> > > > desktop variants, ie. just look at the DRAM rank boundary
> > > > registers to see if both channels have an identical size.
> > > > 
> > > > With this my CL/CTG no longer claim L-shaped memory when I use
> > > > identical sticks. Also tested with non-matching sticks just to
> > > > make sure the L-shaped memory is still properly detected.
> > > > 
> > > > And for completeness let's update the debugfs code to dump
> > > > the correct set of registers on each platform.
> > > > 
> > > > Cc: Chris Wilson <chris@chris-wilson.co.uk>
> > > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > 
> > > Did you check this with the swapping igt? I have some vague memories of
> > > bug reports where somehow the machine was acting like it's L-shaped memory
> > > despite that banks were populated equally. I've iirc tried all kinds of
> > > tricks to figure it out, all to absolutely no avail.
> > 
> > Did you have a specific test in mind? I ran a bunch of things
> > that seemed swizzle related. All passed just fine.
> 
> gem_tiled_swapping should be the one. It tries to cycle your entire system
> memory through tiled buffers into swap and out of it.

Passes with symmetric config, fails with L-shaped config (if I hack
out the L-shape detection of course). So seems pretty solid.

A kernel based self test that looks at the physical address would
still be nice I suppose. Though depending on the size of your RAM
sticks figuring out where exactly the switchover from two channels
to one channels happens probably requires a bit of work due to
the PCI hole/etc.

Both my cl and ctg report this btw:
 bit6 swizzle for X-tiling = bit9/bit10/bit11
 bit6 swizzle for Y-tiling = bit9/bit11
so unfortunately can't be sure the other swizzle modes would be
correctly detected.
Daniel Vetter April 27, 2021, 8:58 a.m. UTC | #6
On Mon, Apr 26, 2021 at 08:18:39PM +0300, Ville Syrjälä wrote:
> On Mon, Apr 26, 2021 at 06:08:59PM +0200, Daniel Vetter wrote:
> > On Thu, Apr 22, 2021 at 04:11:22PM +0300, Ville Syrjälä wrote:
> > > On Thu, Apr 22, 2021 at 11:49:43AM +0200, Daniel Vetter wrote:
> > > > On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> > > > > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > 
> > > > > Currently we try to detect a symmetric memory configurations
> > > > > using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> > > > > either only set on a very specific subset of machines or it
> > > > > just does not exist (it's not mentioned in any public chipset
> > > > > datasheets I've found). As it happens my CL/CTG machines never
> > > > > set said bit, even if I populate the channels with identical
> > > > > sticks.
> > > > > 
> > > > > So let's do the L-shaped memory detection the same way as the
> > > > > desktop variants, ie. just look at the DRAM rank boundary
> > > > > registers to see if both channels have an identical size.
> > > > > 
> > > > > With this my CL/CTG no longer claim L-shaped memory when I use
> > > > > identical sticks. Also tested with non-matching sticks just to
> > > > > make sure the L-shaped memory is still properly detected.
> > > > > 
> > > > > And for completeness let's update the debugfs code to dump
> > > > > the correct set of registers on each platform.
> > > > > 
> > > > > Cc: Chris Wilson <chris@chris-wilson.co.uk>
> > > > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > 
> > > > Did you check this with the swapping igt? I have some vague memories of
> > > > bug reports where somehow the machine was acting like it's L-shaped memory
> > > > despite that banks were populated equally. I've iirc tried all kinds of
> > > > tricks to figure it out, all to absolutely no avail.
> > > 
> > > Did you have a specific test in mind? I ran a bunch of things
> > > that seemed swizzle related. All passed just fine.
> > 
> > gem_tiled_swapping should be the one. It tries to cycle your entire system
> > memory through tiled buffers into swap and out of it.
> 
> Passes with symmetric config, fails with L-shaped config (if I hack
> out the L-shape detection of course). So seems pretty solid.
> 
> A kernel based self test that looks at the physical address would
> still be nice I suppose. Though depending on the size of your RAM
> sticks figuring out where exactly the switchover from two channels
> to one channels happens probably requires a bit of work due to
> the PCI hole/etc.
> 
> Both my cl and ctg report this btw:
>  bit6 swizzle for X-tiling = bit9/bit10/bit11
>  bit6 swizzle for Y-tiling = bit9/bit11
> so unfortunately can't be sure the other swizzle modes would be
> correctly detected.

I think testing-wise this is as good as it gets.
-Daniel
Ville Syrjälä Oct. 4, 2021, 10:36 a.m. UTC | #7
On Tue, Apr 27, 2021 at 10:58:07AM +0200, Daniel Vetter wrote:
> On Mon, Apr 26, 2021 at 08:18:39PM +0300, Ville Syrjälä wrote:
> > On Mon, Apr 26, 2021 at 06:08:59PM +0200, Daniel Vetter wrote:
> > > On Thu, Apr 22, 2021 at 04:11:22PM +0300, Ville Syrjälä wrote:
> > > > On Thu, Apr 22, 2021 at 11:49:43AM +0200, Daniel Vetter wrote:
> > > > > On Wed, Apr 21, 2021 at 06:34:01PM +0300, Ville Syrjala wrote:
> > > > > > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > > 
> > > > > > Currently we try to detect a symmetric memory configurations
> > > > > > using a magic DCC2_MODIFIED_ENHANCED_DISABLE bit. That bit is
> > > > > > either only set on a very specific subset of machines or it
> > > > > > just does not exist (it's not mentioned in any public chipset
> > > > > > datasheets I've found). As it happens my CL/CTG machines never
> > > > > > set said bit, even if I populate the channels with identical
> > > > > > sticks.
> > > > > > 
> > > > > > So let's do the L-shaped memory detection the same way as the
> > > > > > desktop variants, ie. just look at the DRAM rank boundary
> > > > > > registers to see if both channels have an identical size.
> > > > > > 
> > > > > > With this my CL/CTG no longer claim L-shaped memory when I use
> > > > > > identical sticks. Also tested with non-matching sticks just to
> > > > > > make sure the L-shaped memory is still properly detected.
> > > > > > 
> > > > > > And for completeness let's update the debugfs code to dump
> > > > > > the correct set of registers on each platform.
> > > > > > 
> > > > > > Cc: Chris Wilson <chris@chris-wilson.co.uk>
> > > > > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > > > > 
> > > > > Did you check this with the swapping igt? I have some vague memories of
> > > > > bug reports where somehow the machine was acting like it's L-shaped memory
> > > > > despite that banks were populated equally. I've iirc tried all kinds of
> > > > > tricks to figure it out, all to absolutely no avail.
> > > > 
> > > > Did you have a specific test in mind? I ran a bunch of things
> > > > that seemed swizzle related. All passed just fine.
> > > 
> > > gem_tiled_swapping should be the one. It tries to cycle your entire system
> > > memory through tiled buffers into swap and out of it.
> > 
> > Passes with symmetric config, fails with L-shaped config (if I hack
> > out the L-shape detection of course). So seems pretty solid.
> > 
> > A kernel based self test that looks at the physical address would
> > still be nice I suppose. Though depending on the size of your RAM
> > sticks figuring out where exactly the switchover from two channels
> > to one channels happens probably requires a bit of work due to
> > the PCI hole/etc.
> > 
> > Both my cl and ctg report this btw:
> >  bit6 swizzle for X-tiling = bit9/bit10/bit11
> >  bit6 swizzle for Y-tiling = bit9/bit11
> > so unfortunately can't be sure the other swizzle modes would be
> > correctly detected.
> 
> I think testing-wise this is as good as it gets.

So what do we think about putting this in? Currently this only works by
sheer luck more or less. On my machines we have a false positive which
is safe now that the pin quirk got fixed, but if some other machines
have a false negative then things are not going to go so well.
diff mbox series

Patch

diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
index 0fa6c38893f7..754f20768de5 100644
--- a/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
+++ b/drivers/gpu/drm/i915/gt/intel_ggtt_fencing.c
@@ -693,14 +693,15 @@  static void detect_bit_6_swizzle(struct i915_ggtt *ggtt)
 				swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
 				swizzle_y = I915_BIT_6_SWIZZLE_9_17;
 			}
-			break;
-		}
 
-		/* check for L-shaped memory aka modified enhanced addressing */
-		if (IS_GEN(i915, 4) &&
-		    !(intel_uncore_read(uncore, DCC2) & DCC2_MODIFIED_ENHANCED_DISABLE)) {
-			swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
-			swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
+			/* check for L-shaped memory aka modified enhanced addressing */
+			if (IS_GEN(i915, 4) &&
+			    intel_uncore_read16(uncore, C0DRB3_CL) !=
+			    intel_uncore_read16(uncore, C1DRB3_CL)) {
+				swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
+				swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
+			}
+			break;
 		}
 
 		if (dcc == 0xffffffff) {
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 8dd374691102..6de11ffcde38 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -621,10 +621,18 @@  static int i915_swizzle_info(struct seq_file *m, void *data)
 			   intel_uncore_read(uncore, DCC));
 		seq_printf(m, "DDC2 = 0x%08x\n",
 			   intel_uncore_read(uncore, DCC2));
-		seq_printf(m, "C0DRB3 = 0x%04x\n",
-			   intel_uncore_read16(uncore, C0DRB3_BW));
-		seq_printf(m, "C1DRB3 = 0x%04x\n",
-			   intel_uncore_read16(uncore, C1DRB3_BW));
+
+		if (IS_G45(dev_priv) || IS_I965G(dev_priv) || IS_G33(dev_priv)) {
+			seq_printf(m, "C0DRB3 = 0x%04x\n",
+				   intel_uncore_read16(uncore, C0DRB3_BW));
+			seq_printf(m, "C1DRB3 = 0x%04x\n",
+				   intel_uncore_read16(uncore, C1DRB3_BW));
+		} else if (IS_GEN(dev_priv, 4)) {
+			seq_printf(m, "C0DRB3 = 0x%04x\n",
+				   intel_uncore_read16(uncore, C0DRB3_CL));
+			seq_printf(m, "C1DRB3 = 0x%04x\n",
+				   intel_uncore_read16(uncore, C1DRB3_CL));
+		}
 	} else if (INTEL_GEN(dev_priv) >= 6) {
 		seq_printf(m, "MAD_DIMM_C0 = 0x%08x\n",
 			   intel_uncore_read(uncore, MAD_DIMM_C0));
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 0587b2455ea1..055c258179a1 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -3790,6 +3790,10 @@  static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define C0DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x206)
 #define C1DRB3_BW		_MMIO(MCHBAR_MIRROR_BASE + 0x606)
 
+/* 965gm,ctg DRAM channel configuration */
+#define C0DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1206)
+#define C1DRB3_CL		_MMIO(MCHBAR_MIRROR_BASE + 0x1306)
+
 /* snb MCH registers for reading the DRAM channel configuration */
 #define MAD_DIMM_C0			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5004)
 #define MAD_DIMM_C1			_MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5008)