From: "Ville Syrjälä" <ville.syrjala@linux.intel.com>
To: "Sharma, Swati2" <swati2.sharma@intel.com>
Cc: jani.nikula@intel.com, daniel.vetter@ffwll.ch,
intel-gfx@lists.freedesktop.org, ankit.k.nautiyal@intel.com
Subject: Re: [v5][PATCH 11/11] drm/i915: Add intel_color_lut_equal() to compare hw and sw gamma/degamma lut values
Date: Tue, 7 May 2019 13:53:08 +0300 [thread overview]
Message-ID: <20190507105308.GJ24299@intel.com> (raw)
In-Reply-To: <feca5839-0cf2-41cd-51a6-77b385ece7d8@intel.com>
On Tue, May 07, 2019 at 12:27:07AM +0530, Sharma, Swati2 wrote:
> On 07-May-19 12:03 AM, Ville Syrjälä wrote:
>
> > On Sat, May 04, 2019 at 10:41:40PM +0530, Swati Sharma wrote:
> >> v3: Rebase
> >> v4: -Renamed intel_compare_color_lut() to intel_color_lut_equal() [Jani]
> >> -Added the default label above the correct label [Jani]
> >> -Corrected smatch warn "variable dereferenced before check" [Dan Carpenter]
> >> v5: -Added condition (!blob1 && !blob2) return true [Jani]
> >> -Called PIPE_CONF_CHECK_COLOR_LUT inside if (!adjust) [Jani]
> >> -Added #undef PIPE_CONF_CHECK_COLOR_LUT [Jani]
> >>
> >> There are few things wrong in this patch:
> >> [1] For chv bit precision is 14, on what basis it should be assigned?
> > Like everything else it will more or less be a reverse of the
> > compute side. For CHV we need to look at cgm_mode, gamma_enable,
> > gamma_mode, and c8_planes. Hmm. We actually don't have readout for
> > c8_planes so I guess we'll have to make some kind of exception for
> > that one.
>
> By this I meant was, since I am assigning bit_precision on the basis
> of gamma_mode in the compare function like
> + case GAMMA_MODE_MODE_8BIT:
> + bit_precision = 8;
> etc. We have 8BIT, 10BIT and 12BIT GAMMA_MODES only.How will I
> assign 14BIT on the basis of GAMMA_MODES (or) is there some other
> option to assign precision. Please see the comparison code below.
Yeah, gamma_mode by itself is not sufficient in some of the cases.
It's highly platform dependent how you figure out the precision.
>
> >
> >> [2] For glk and icl, degamma LUT values are not rounded off, there
> >> should err=0 if using same function, how to make that exception?
> > You mean the degamma? Just set precision==16? Maybe I'm not
> > understanding the question.
>
> Again same query as above, if I set precision as 16..on what basis?
> Which GAMMA_MODE? (or) default i should set as 16?
The glk+ degamma is always 16bpc, so you just hardcode it.
>
> >
> >> [3] For glk, bit precision is 10 but gamma mode is 8BIT?
> > Not sure what you mean. glk gamma_mode works just like with
> > other ilk+ platforms (apart from not having the csc_mode
> > gamma vs. degamma control).
>
> Sorry..my bad!
>
> >
> > I suspect the most annoying case is ivb-bdw 10bit gamma mode
> > since we probably don't have the full readout to do the reverse
> > mapping of whether the LUT is used as gamma or degamma. But I
> > guess we'll just have to compare it against whichever one the
> > software state has.
>
> One last query, as there is difference in precision of gamma and
> degamma..we have one comparison function..how will we get to know whether
> blob received is degamma or gamma? Do we need to pass some kind of enum value
> to know comparison needs to be done for gamma or degamma?
I think the comparison should just compare two blobs. Doesn't matter
what they are really.
>
> >
> >> Signed-off-by: Swati Sharma <swati2.sharma@intel.com>
> >> ---
> >> drivers/gpu/drm/i915/intel_color.c | 54 ++++++++++++++++++++++++++++++++++++
> >> drivers/gpu/drm/i915/intel_color.h | 6 ++++
> >> drivers/gpu/drm/i915/intel_display.c | 12 ++++++++
> >> 3 files changed, 72 insertions(+)
> >>
> >> diff --git a/drivers/gpu/drm/i915/intel_color.c b/drivers/gpu/drm/i915/intel_color.c
> >> index 695b76d..73cb901 100644
> >> --- a/drivers/gpu/drm/i915/intel_color.c
> >> +++ b/drivers/gpu/drm/i915/intel_color.c
> >> @@ -1630,6 +1630,60 @@ static void ilk_read_luts(struct intel_crtc_state *crtc_state)
> >> crtc_state->base.gamma_lut = ilk_read_gamma_lut(crtc_state);
> >> }
> >>
> >> +static inline bool err_check(struct drm_color_lut *sw_lut, struct drm_color_lut *hw_lut, u32 err)
> >> +{
> >> + return ((abs((long)hw_lut->red - sw_lut->red)) <= err) &&
> >> + ((abs((long)hw_lut->blue - sw_lut->blue)) <= err) &&
> >> + ((abs((long)hw_lut->green - sw_lut->green)) <= err);
> >> +}
> >> +
> >> +bool intel_color_lut_equal(struct drm_property_blob *blob1,
> >> + struct drm_property_blob *blob2,
> >> + u32 gamma_mode)
> >> +{
> >> + struct drm_color_lut *sw_lut, *hw_lut;
> >> + int sw_lut_size, hw_lut_size, i;
> >> + u32 bit_precision, err;
> >> +
> >> + if (!blob1 && !blob2)
> >> + return true;
> >> +
> >> + if (!blob1 || !blob2)
> >> + return false;
> >> +
> >> + sw_lut_size = drm_color_lut_size(blob1);
> >> + hw_lut_size = drm_color_lut_size(blob2);
> >> +
> >> + if (sw_lut_size != hw_lut_size)
> >> + return false;
> >> +
> >> + switch(gamma_mode) {
> >> + default:
> >> + case GAMMA_MODE_MODE_8BIT:
> >> + bit_precision = 8;
> >> + break;
> >> + case GAMMA_MODE_MODE_10BIT:
> >> + case GAMMA_MODE_MODE_SPLIT:
> >> + bit_precision = 10;
> >> + break;
> >> + case GAMMA_MODE_MODE_12BIT:
> >> + bit_precision = 12;
> >> + break;
> >> + }
> >> +
> >> + sw_lut = blob1->data;
> >> + hw_lut = blob2->data;
> >> +
> >> + err = 0xffff >> bit_precision;
> >> +
> >> + for (i = 0; i < sw_lut_size; i++) {
> >> + if (!err_check(&hw_lut[i], &sw_lut[i], err))
> >> + return false;
> >> + }
> >> +
> >> + return true;
> >> +}
> >> +
> >> void intel_color_init(struct intel_crtc *crtc)
> >> {
> >> struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
> >> diff --git a/drivers/gpu/drm/i915/intel_color.h b/drivers/gpu/drm/i915/intel_color.h
> >> index fc53de9..b525c80 100644
> >> --- a/drivers/gpu/drm/i915/intel_color.h
> >> +++ b/drivers/gpu/drm/i915/intel_color.h
> >> @@ -6,13 +6,19 @@
> >> #ifndef __INTEL_COLOR_H__
> >> #define __INTEL_COLOR_H__
> >>
> >> +#include <linux/types.h>
> >> +
> >> struct intel_crtc_state;
> >> struct intel_crtc;
> >> +struct drm_property_blob;
> >>
> >> void intel_color_init(struct intel_crtc *crtc);
> >> int intel_color_check(struct intel_crtc_state *crtc_state);
> >> void intel_color_commit(const struct intel_crtc_state *crtc_state);
> >> void intel_color_load_luts(const struct intel_crtc_state *crtc_state);
> >> void intel_color_read_luts(struct intel_crtc_state *crtc_state);
> >> +bool intel_color_lut_equal(struct drm_property_blob *blob1,
> >> + struct drm_property_blob *blob2,
> >> + u32 gamma_mode);
> >>
> >> #endif /* __INTEL_COLOR_H__ */
> >> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> >> index 791974b..a713171 100644
> >> --- a/drivers/gpu/drm/i915/intel_display.c
> >> +++ b/drivers/gpu/drm/i915/intel_display.c
> >> @@ -12287,6 +12287,14 @@ static bool fastboot_enabled(struct drm_i915_private *dev_priv)
> >> } \
> >> } while (0)
> >>
> >> +#define PIPE_CONF_CHECK_COLOR_LUT(name, gamma_mode) do { \
> >> + if (!intel_color_lut_equal(current_config->name, pipe_config->name, gamma_mode)) { \
> >> + pipe_config_err(adjust, __stringify(name), \
> >> + "hw_state doesn't match sw_state\n"); \
> >> + ret = false; \
> >> + } \
> >> +} while (0)
> >> +
> >> #define PIPE_CONF_QUIRK(quirk) \
> >> ((current_config->quirks | pipe_config->quirks) & (quirk))
> >>
> >> @@ -12376,6 +12384,9 @@ static bool fastboot_enabled(struct drm_i915_private *dev_priv)
> >> PIPE_CONF_CHECK_X(csc_mode);
> >> PIPE_CONF_CHECK_BOOL(gamma_enable);
> >> PIPE_CONF_CHECK_BOOL(csc_enable);
> >> +
> >> + PIPE_CONF_CHECK_COLOR_LUT(base.gamma_lut, pipe_config->gamma_mode);
> >> + PIPE_CONF_CHECK_COLOR_LUT(base.degamma_lut, pipe_config->gamma_mode);
> >> }
> >>
> >> PIPE_CONF_CHECK_BOOL(double_wide);
> >> @@ -12438,6 +12449,7 @@ static bool fastboot_enabled(struct drm_i915_private *dev_priv)
> >> #undef PIPE_CONF_CHECK_FLAGS
> >> #undef PIPE_CONF_CHECK_CLOCK_FUZZY
> >> #undef PIPE_CONF_QUIRK
> >> +#undef PIPE_CONF_CHECK_COLOR_LUT
> >>
> >> return ret;
> >> }
> >> --
> >> 1.9.1
>
> --
> ~Swati Sharma
>
--
Ville Syrjälä
Intel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2019-05-07 10:53 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-05-04 17:11 [PATCH 00/11] drm/i915: adding state checker for gamma lut values Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 01/11] drm/i915: Introduce vfunc read_luts() to create hw lut Swati Sharma
2019-05-06 13:21 ` Jani Nikula
2019-05-06 13:29 ` Ville Syrjälä
2019-05-06 13:41 ` Jani Nikula
2019-05-06 14:24 ` Sharma, Swati2
2019-05-07 6:48 ` Jani Nikula
2019-05-07 7:08 ` Jani Nikula
2019-05-04 17:11 ` [v5][PATCH 02/11] drm/i915: Enable intel_color_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 03/11] drm/i915: Extract i9xx_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 04/11] drm/i915: Extract chv_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 05/11] drm/i915: Extract i965_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 06/11] drm/i915: Extract icl_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 07/11] drm/i915: Extract glk_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 08/11] drm/i915: Extract bdw_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 09/11] drm/i915: Extract ivb_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 10/11] drm/i915: Extract ilk_read_luts() Swati Sharma
2019-05-04 17:11 ` [v5][PATCH 11/11] drm/i915: Add intel_color_lut_equal() to compare hw and sw gamma/degamma lut values Swati Sharma
2019-05-06 18:33 ` Ville Syrjälä
2019-05-06 18:57 ` Sharma, Swati2
2019-05-07 10:53 ` Ville Syrjälä [this message]
2019-05-04 17:26 ` ✗ Fi.CI.CHECKPATCH: warning for drm/i915: adding state checker for gamma lut values (rev7) Patchwork
2019-05-04 17:31 ` ✗ Fi.CI.SPARSE: " Patchwork
2019-05-04 17:40 ` ✗ Fi.CI.BAT: failure " Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190507105308.GJ24299@intel.com \
--to=ville.syrjala@linux.intel.com \
--cc=ankit.k.nautiyal@intel.com \
--cc=daniel.vetter@ffwll.ch \
--cc=intel-gfx@lists.freedesktop.org \
--cc=jani.nikula@intel.com \
--cc=swati2.sharma@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox