[Intel-gfx] [PATCH v6 08/10] drm/i915/dsb: Enable gamma lut programming using DSB.
Animesh Manna
animesh.manna at intel.com
Tue Sep 17 09:19:57 UTC 2019
On 9/17/2019 1:00 PM, Jani Nikula wrote:
> On Thu, 12 Sep 2019, Animesh Manna <animesh.manna at intel.com> wrote:
>> On 9/12/2019 6:37 PM, Jani Nikula wrote:
>>> On Thu, 12 Sep 2019, Animesh Manna <animesh.manna at intel.com> wrote:
>>>> Gamma lut programming can be programmed using DSB
>>>> where bulk register programming can be done using indexed
>>>> register write which takes number of data and the mmio offset
>>>> to be written.
>>>>
>>>> Currently enabled for 12-bit gamma LUT which is enabled by
>>>> default and later 8-bit/10-bit will be enabled in future
>>>> based on need.
>>>>
>>>> v1: Initial version.
>>>> v2: Directly call dsb-api at callsites. (Jani)
>>>> v3:
>>>> - modified the code as per single dsb instance per crtc. (Shashank)
>>>> - Added dsb get/put call in platform specific load_lut hook. (Jani)
>>>> - removed dsb pointer from dev_priv. (Jani)
>>>> v4: simplified code by dropping ref-count implementation. (Shashank)
>>>>
>>>> Cc: Jani Nikula <jani.nikula at intel.com>
>>>> Cc: Rodrigo Vivi <rodrigo.vivi at intel.com>
>>>> Cc: Shashank Sharma <shashank.sharma at intel.com>
>>>> Signed-off-by: Animesh Manna <animesh.manna at intel.com>
>>>> ---
>>>> drivers/gpu/drm/i915/display/intel_color.c | 57 +++++++++++++---------
>>>> 1 file changed, 35 insertions(+), 22 deletions(-)
>>>>
>>>> diff --git a/drivers/gpu/drm/i915/display/intel_color.c b/drivers/gpu/drm/i915/display/intel_color.c
>>>> index 318308dc136c..b6b9f0e5166b 100644
>>>> --- a/drivers/gpu/drm/i915/display/intel_color.c
>>>> +++ b/drivers/gpu/drm/i915/display/intel_color.c
>>>> @@ -611,12 +611,13 @@ static void bdw_load_lut_10(struct intel_crtc *crtc,
>>>> static void ivb_load_lut_ext_max(struct intel_crtc *crtc)
>>>> {
>>>> struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
>>>> + struct intel_dsb *dsb = intel_dsb_get(crtc);
>>> When you've done enough kernel programming, you learn to expect get/put
>>> to be paired, and assume it's a bug if this isn't the case.
>> Already have done it in my previous version,
>> https://patchwork.freedesktop.org/patch/329481/?series=63013&rev=5
>> https://patchwork.freedesktop.org/patch/329482/?series=63013&rev=5
>>
>> Can you please suggest if it is ok?
> Yes, but you need to get it right. It seems that the inc and dec can get
> out of sync. Also no warnings against putting too many times.
Hi Jani,
Extra put will not decrement as it is under cmd_buf condition check.
Once it is zero cmd_buf will be assigned to NULL.
if (dsb->cmd_buf) {
atomic_dec(&dsb->refcount);
if (!atomic_read(&dsb->refcount)) {
mutex_lock(&i915->drm.struct_mutex);
i915_gem_object_unpin_map(dsb->vma->obj);
i915_vma_unpin_and_release(&dsb->vma, 0);
dsb->cmd_buf = NULL;
mutex_unlock(&i915->drm.struct_mutex);
}
}
Do you want me to add some warning if refcount is already zero and user
called put?
Can you please help me to understand where inc and dec can go out of sync.
Regards,
Animesh
>
> BR,
> Jani.
>
>
>
>> Regards,
>> Animesh
>>> So yeah, I did say it's okay not to have refcounting at first, *but* the
>>> expectation that get/put go in pairs is so deeply ingrained that I
>>> didn't even think to say you shouldn't have this kind of asymmetry.
>>>
>>> So this creates a problem, how do we pass the dsb pointer here, as
>>> without refcounting you can't actually have the put here as well,
>>> because you throw the stuff out before the commit.
>>>
>>> Maybe the easy answer is that we should just do the get and put at
>>> intel_crtc_init and intel_crtc_destroy. Or we could do it at atomic
>>> check and free.
>>>
>>> Ville, which approach would conflict with your future vblank worker
>>> stuff the least?
>>>
>>>
>>> BR,
>>> Jani.
>>>
>>>
>>>
>>>> enum pipe pipe = crtc->pipe;
>>>>
>>>> /* Program the max register to clamp values > 1.0. */
>>>> - I915_WRITE(PREC_PAL_EXT_GC_MAX(pipe, 0), 1 << 16);
>>>> - I915_WRITE(PREC_PAL_EXT_GC_MAX(pipe, 1), 1 << 16);
>>>> - I915_WRITE(PREC_PAL_EXT_GC_MAX(pipe, 2), 1 << 16);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_EXT_GC_MAX(pipe, 0), 1 << 16);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_EXT_GC_MAX(pipe, 1), 1 << 16);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_EXT_GC_MAX(pipe, 2), 1 << 16);
>>>>
>>>> /*
>>>> * Program the gc max 2 register to clamp values > 1.0.
>>>> @@ -624,9 +625,12 @@ static void ivb_load_lut_ext_max(struct intel_crtc *crtc)
>>>> * from 3.0 to 7.0
>>>> */
>>>> if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
>>>> - I915_WRITE(PREC_PAL_EXT2_GC_MAX(pipe, 0), 1 << 16);
>>>> - I915_WRITE(PREC_PAL_EXT2_GC_MAX(pipe, 1), 1 << 16);
>>>> - I915_WRITE(PREC_PAL_EXT2_GC_MAX(pipe, 2), 1 << 16);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_EXT2_GC_MAX(pipe, 0),
>>>> + 1 << 16);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_EXT2_GC_MAX(pipe, 1),
>>>> + 1 << 16);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_EXT2_GC_MAX(pipe, 2),
>>>> + 1 << 16);
>>>> }
>>>> }
>>>>
>>>> @@ -787,22 +791,22 @@ icl_load_gcmax(const struct intel_crtc_state *crtc_state,
>>>> const struct drm_color_lut *color)
>>>> {
>>>> struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
>>>> - struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
>>>> + struct intel_dsb *dsb = intel_dsb_get(crtc);
>>>> enum pipe pipe = crtc->pipe;
>>>>
>>>> /* Fixme: LUT entries are 16 bit only, so we can prog 0xFFFF max */
>>>> - I915_WRITE(PREC_PAL_GC_MAX(pipe, 0), color->red);
>>>> - I915_WRITE(PREC_PAL_GC_MAX(pipe, 1), color->green);
>>>> - I915_WRITE(PREC_PAL_GC_MAX(pipe, 2), color->blue);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_GC_MAX(pipe, 0), color->red);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_GC_MAX(pipe, 1), color->green);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_GC_MAX(pipe, 2), color->blue);
>>>> }
>>>>
>>>> static void
>>>> icl_program_gamma_superfine_segment(const struct intel_crtc_state *crtc_state)
>>>> {
>>>> struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
>>>> - struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
>>>> const struct drm_property_blob *blob = crtc_state->base.gamma_lut;
>>>> const struct drm_color_lut *lut = blob->data;
>>>> + struct intel_dsb *dsb = intel_dsb_get(crtc);
>>>> enum pipe pipe = crtc->pipe;
>>>> u32 i;
>>>>
>>>> @@ -813,15 +817,16 @@ icl_program_gamma_superfine_segment(const struct intel_crtc_state *crtc_state)
>>>> * Superfine segment has 9 entries, corresponding to values
>>>> * 0, 1/(8 * 128 * 256), 2/(8 * 128 * 256) .... 8/(8 * 128 * 256).
>>>> */
>>>> - I915_WRITE(PREC_PAL_MULTI_SEG_INDEX(pipe), PAL_PREC_AUTO_INCREMENT);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_MULTI_SEG_INDEX(pipe),
>>>> + PAL_PREC_AUTO_INCREMENT);
>>>>
>>>> for (i = 0; i < 9; i++) {
>>>> const struct drm_color_lut *entry = &lut[i];
>>>>
>>>> - I915_WRITE(PREC_PAL_MULTI_SEG_DATA(pipe),
>>>> - ilk_lut_12p4_ldw(entry));
>>>> - I915_WRITE(PREC_PAL_MULTI_SEG_DATA(pipe),
>>>> - ilk_lut_12p4_udw(entry));
>>>> + intel_dsb_indexed_reg_write(dsb, PREC_PAL_MULTI_SEG_DATA(pipe),
>>>> + ilk_lut_12p4_ldw(entry));
>>>> + intel_dsb_indexed_reg_write(dsb, PREC_PAL_MULTI_SEG_DATA(pipe),
>>>> + ilk_lut_12p4_udw(entry));
>>>> }
>>>> }
>>>>
>>>> @@ -829,10 +834,10 @@ static void
>>>> icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
>>>> {
>>>> struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
>>>> - struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
>>>> const struct drm_property_blob *blob = crtc_state->base.gamma_lut;
>>>> const struct drm_color_lut *lut = blob->data;
>>>> const struct drm_color_lut *entry;
>>>> + struct intel_dsb *dsb = intel_dsb_get(crtc);
>>>> enum pipe pipe = crtc->pipe;
>>>> u32 i;
>>>>
>>>> @@ -847,11 +852,13 @@ icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
>>>> * PAL_PREC_INDEX[0] and PAL_PREC_INDEX[1] map to seg2[1],
>>>> * with seg2[0] being unused by the hardware.
>>>> */
>>>> - I915_WRITE(PREC_PAL_INDEX(pipe), PAL_PREC_AUTO_INCREMENT);
>>>> + intel_dsb_reg_write(dsb, PREC_PAL_INDEX(pipe), PAL_PREC_AUTO_INCREMENT);
>>>> for (i = 1; i < 257; i++) {
>>>> entry = &lut[i * 8];
>>>> - I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_ldw(entry));
>>>> - I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_udw(entry));
>>>> + intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
>>>> + ilk_lut_12p4_ldw(entry));
>>>> + intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
>>>> + ilk_lut_12p4_udw(entry));
>>>> }
>>>>
>>>> /*
>>>> @@ -868,8 +875,10 @@ icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
>>>> */
>>>> for (i = 0; i < 256; i++) {
>>>> entry = &lut[i * 8 * 128];
>>>> - I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_ldw(entry));
>>>> - I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_udw(entry));
>>>> + intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
>>>> + ilk_lut_12p4_ldw(entry));
>>>> + intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
>>>> + ilk_lut_12p4_udw(entry));
>>>> }
>>>>
>>>> /* The last entry in the LUT is to be programmed in GCMAX */
>>>> @@ -882,6 +891,7 @@ static void icl_load_luts(const struct intel_crtc_state *crtc_state)
>>>> {
>>>> const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
>>>> struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
>>>> + struct intel_dsb *dsb = intel_dsb_get(crtc);
>>>>
>>>> if (crtc_state->base.degamma_lut)
>>>> glk_load_degamma_lut(crtc_state);
>>>> @@ -900,6 +910,9 @@ static void icl_load_luts(const struct intel_crtc_state *crtc_state)
>>>> bdw_load_lut_10(crtc, gamma_lut, PAL_PREC_INDEX_VALUE(0));
>>>> ivb_load_lut_ext_max(crtc);
>>>> }
>>>> +
>>>> + intel_dsb_commit(dsb);
>>>> + intel_dsb_put(dsb);
>>>> }
>>>>
>>>> static u32 chv_cgm_degamma_ldw(const struct drm_color_lut *color)
More information about the Intel-gfx
mailing list