[Intel-gfx] [PATCH v2 1/4] drm/i915/fbc: Parametrize FBC register offsets
Jani Nikula
jani.nikula at intel.com
Tue Dec 14 18:34:31 UTC 2021
On Tue, 14 Dec 2021, Ville Syrjälä <ville.syrjala at linux.intel.com> wrote:
> On Mon, Dec 13, 2021 at 09:54:04PM +0200, Jani Nikula wrote:
>> On Mon, 13 Dec 2021, Ville Syrjala <ville.syrjala at linux.intel.com> wrote:
>> > From: Ville Syrjälä <ville.syrjala at linux.intel.com>
>> >
>> > Parametrize ilk+ FBC register offsets based on the FBC instance.
>> >
>> > v2: More intel_ namespace (Jani)
>> >
>> > Cc: Jani Nikula <jani.nikula at intel.com>
>> > Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
>>
>> Some questions below, apart from that,
>>
>> Reviewed-by: Jani Nikula <jani.nikula at intel.com>
>>
>> > ---
>> > drivers/gpu/drm/i915/display/intel_fbc.c | 34 +++++++++++++-----------
>> > drivers/gpu/drm/i915/display/intel_fbc.h | 6 +++++
>> > drivers/gpu/drm/i915/i915_reg.h | 34 ++++++++++++------------
>> > drivers/gpu/drm/i915/intel_pm.c | 31 ++++++++++++---------
>> > 4 files changed, 60 insertions(+), 45 deletions(-)
>> >
>> > diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c
>> > index 8be01b93015f..112aafa72253 100644
>> > --- a/drivers/gpu/drm/i915/display/intel_fbc.c
>> > +++ b/drivers/gpu/drm/i915/display/intel_fbc.c
>> > @@ -85,6 +85,8 @@ struct intel_fbc {
>> > struct drm_mm_node compressed_fb;
>> > struct drm_mm_node compressed_llb;
>> >
>> > + enum intel_fbc_id id;
>> > +
>> > u8 limit;
>> >
>> > bool false_color;
>> > @@ -454,10 +456,10 @@ static void ilk_fbc_activate(struct intel_fbc *fbc)
>> > struct intel_fbc_state *fbc_state = &fbc->state;
>> > struct drm_i915_private *i915 = fbc->i915;
>> >
>> > - intel_de_write(i915, ILK_DPFC_FENCE_YOFF,
>> > + intel_de_write(i915, ILK_DPFC_FENCE_YOFF(fbc->id),
>> > fbc_state->fence_y_offset);
>> >
>> > - intel_de_write(i915, ILK_DPFC_CONTROL,
>> > + intel_de_write(i915, ILK_DPFC_CONTROL(fbc->id),
>> > DPFC_CTL_EN | g4x_dpfc_ctl(fbc));
>> > }
>> >
>> > @@ -467,28 +469,28 @@ static void ilk_fbc_deactivate(struct intel_fbc *fbc)
>> > u32 dpfc_ctl;
>> >
>> > /* Disable compression */
>> > - dpfc_ctl = intel_de_read(i915, ILK_DPFC_CONTROL);
>> > + dpfc_ctl = intel_de_read(i915, ILK_DPFC_CONTROL(fbc->id));
>> > if (dpfc_ctl & DPFC_CTL_EN) {
>> > dpfc_ctl &= ~DPFC_CTL_EN;
>> > - intel_de_write(i915, ILK_DPFC_CONTROL, dpfc_ctl);
>> > + intel_de_write(i915, ILK_DPFC_CONTROL(fbc->id), dpfc_ctl);
>> > }
>> > }
>> >
>> > static bool ilk_fbc_is_active(struct intel_fbc *fbc)
>> > {
>> > - return intel_de_read(fbc->i915, ILK_DPFC_CONTROL) & DPFC_CTL_EN;
>> > + return intel_de_read(fbc->i915, ILK_DPFC_CONTROL(fbc->id)) & DPFC_CTL_EN;
>> > }
>> >
>> > static bool ilk_fbc_is_compressing(struct intel_fbc *fbc)
>> > {
>> > - return intel_de_read(fbc->i915, ILK_DPFC_STATUS) & DPFC_COMP_SEG_MASK;
>> > + return intel_de_read(fbc->i915, ILK_DPFC_STATUS(fbc->id)) & DPFC_COMP_SEG_MASK;
>> > }
>> >
>> > static void ilk_fbc_program_cfb(struct intel_fbc *fbc)
>> > {
>> > struct drm_i915_private *i915 = fbc->i915;
>> >
>> > - intel_de_write(i915, ILK_DPFC_CB_BASE, fbc->compressed_fb.start);
>> > + intel_de_write(i915, ILK_DPFC_CB_BASE(fbc->id), fbc->compressed_fb.start);
>> > }
>> >
>> > static const struct intel_fbc_funcs ilk_fbc_funcs = {
>> > @@ -524,8 +526,8 @@ static void snb_fbc_nuke(struct intel_fbc *fbc)
>> > {
>> > struct drm_i915_private *i915 = fbc->i915;
>> >
>> > - intel_de_write(i915, MSG_FBC_REND_STATE, FBC_REND_NUKE);
>> > - intel_de_posting_read(i915, MSG_FBC_REND_STATE);
>> > + intel_de_write(i915, MSG_FBC_REND_STATE(fbc->id), FBC_REND_NUKE);
>> > + intel_de_posting_read(i915, MSG_FBC_REND_STATE(fbc->id));
>> > }
>> >
>> > static const struct intel_fbc_funcs snb_fbc_funcs = {
>> > @@ -547,7 +549,7 @@ static void glk_fbc_program_cfb_stride(struct intel_fbc *fbc)
>> > val |= FBC_STRIDE_OVERRIDE |
>> > FBC_STRIDE(fbc_state->override_cfb_stride / fbc->limit);
>> >
>> > - intel_de_write(i915, GLK_FBC_STRIDE, val);
>> > + intel_de_write(i915, GLK_FBC_STRIDE(fbc->id), val);
>> > }
>> >
>> > static void skl_fbc_program_cfb_stride(struct intel_fbc *fbc)
>> > @@ -598,19 +600,19 @@ static void ivb_fbc_activate(struct intel_fbc *fbc)
>> > if (i915->ggtt.num_fences)
>> > snb_fbc_program_fence(fbc);
>> >
>> > - intel_de_write(i915, ILK_DPFC_CONTROL,
>> > + intel_de_write(i915, ILK_DPFC_CONTROL(fbc->id),
>> > DPFC_CTL_EN | ivb_dpfc_ctl(fbc));
>> > }
>> >
>> > static bool ivb_fbc_is_compressing(struct intel_fbc *fbc)
>> > {
>> > - return intel_de_read(fbc->i915, ILK_DPFC_STATUS2) & DPFC_COMP_SEG_MASK_IVB;
>> > + return intel_de_read(fbc->i915, ILK_DPFC_STATUS2(fbc->id)) & DPFC_COMP_SEG_MASK_IVB;
>> > }
>> >
>> > static void ivb_fbc_set_false_color(struct intel_fbc *fbc,
>> > bool enable)
>> > {
>> > - intel_de_rmw(fbc->i915, ILK_DPFC_CONTROL,
>> > + intel_de_rmw(fbc->i915, ILK_DPFC_CONTROL(fbc->id),
>> > DPFC_CTL_FALSE_COLOR, enable ? DPFC_CTL_FALSE_COLOR : 0);
>> > }
>> >
>> > @@ -1620,7 +1622,8 @@ void intel_fbc_add_plane(struct intel_fbc *fbc, struct intel_plane *plane)
>> > fbc->possible_framebuffer_bits |= plane->frontbuffer_bit;
>> > }
>> >
>> > -static struct intel_fbc *intel_fbc_create(struct drm_i915_private *i915)
>> > +static struct intel_fbc *intel_fbc_create(struct drm_i915_private *i915,
>> > + enum intel_fbc_id fbc_id)
>> > {
>> > struct intel_fbc *fbc;
>> >
>> > @@ -1628,6 +1631,7 @@ static struct intel_fbc *intel_fbc_create(struct drm_i915_private *i915)
>> > if (!fbc)
>> > return NULL;
>> >
>> > + fbc->id = fbc_id;
>> > fbc->i915 = i915;
>> > INIT_WORK(&fbc->underrun_work, intel_fbc_underrun_work_fn);
>> > mutex_init(&fbc->lock);
>> > @@ -1671,7 +1675,7 @@ void intel_fbc_init(struct drm_i915_private *i915)
>> > if (!HAS_FBC(i915))
>> > return;
>> >
>> > - fbc = intel_fbc_create(i915);
>> > + fbc = intel_fbc_create(i915, INTEL_FBC_A);
>> > if (!fbc)
>> > return;
>> >
>> > diff --git a/drivers/gpu/drm/i915/display/intel_fbc.h b/drivers/gpu/drm/i915/display/intel_fbc.h
>> > index 07ad0411fcc3..7b7631aec527 100644
>> > --- a/drivers/gpu/drm/i915/display/intel_fbc.h
>> > +++ b/drivers/gpu/drm/i915/display/intel_fbc.h
>> > @@ -17,6 +17,12 @@ struct intel_fbc;
>> > struct intel_plane;
>> > struct intel_plane_state;
>> >
>> > +enum intel_fbc_id {
>> > + INTEL_FBC_A,
>> > +
>> > + I915_MAX_FBCS,
>> > +};
>> > +
>> > int intel_fbc_atomic_check(struct intel_atomic_state *state);
>> > bool intel_fbc_pre_update(struct intel_atomic_state *state,
>> > struct intel_crtc *crtc);
>> > diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
>> > index d27ba273cc68..698a023e70f5 100644
>> > --- a/drivers/gpu/drm/i915/i915_reg.h
>> > +++ b/drivers/gpu/drm/i915/i915_reg.h
>> > @@ -3386,10 +3386,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
>> > #define FBC_LL_SIZE (1536)
>> >
>> > /* Framebuffer compression for GM45+ */
>> > -#define DPFC_CB_BASE _MMIO(0x3200)
>> > -#define ILK_DPFC_CB_BASE _MMIO(0x43200)
>> > -#define DPFC_CONTROL _MMIO(0x3208)
>> > -#define ILK_DPFC_CONTROL _MMIO(0x43208)
>> > +#define DPFC_CB_BASE _MMIO(0x3200)
>> > +#define ILK_DPFC_CB_BASE(fbc_id) _MMIO_PIPE((fbc_id), 0x43200, 0x43240)
>> > +#define DPFC_CONTROL _MMIO(0x3208)
>> > +#define ILK_DPFC_CONTROL(fbc_id) _MMIO_PIPE((fbc_id), 0x43208, 0x43248)
>> > #define DPFC_CTL_EN REG_BIT(31)
>> > #define DPFC_CTL_PLANE_MASK_G4X REG_BIT(30) /* g4x-snb */
>> > #define DPFC_CTL_PLANE_G4X(i9xx_plane) REG_FIELD_PREP(DPFC_CTL_PLANE_MASK_G4X, (i9xx_plane))
>> > @@ -3407,28 +3407,28 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
>> > #define DPFC_CTL_LIMIT_4X REG_FIELD_PREP(DPFC_CTL_LIMIT_MASK, 2)
>> > #define DPFC_CTL_FENCENO_MASK REG_GENMASK(3, 0)
>> > #define DPFC_CTL_FENCENO(fence) REG_FIELD_PREP(DPFC_CTL_FENCENO_MASK, (fence))
>> > -#define DPFC_RECOMP_CTL _MMIO(0x320c)
>> > -#define ILK_DPFC_RECOMP_CTL _MMIO(0x4320c)
>> > +#define DPFC_RECOMP_CTL _MMIO(0x320c)
>> > +#define ILK_DPFC_RECOMP_CTL(fbc_id) _MMIO_PIPE((fbc_id), 0x4320c, 0x4324c)
>>
>> This is display 5 and 6 only, right?
>
> Hmm. Looks like it may be removed in adl-p. But definitely still
> present in tgl.
>
>> Will there be a register instance
>> for fbc_id > INTEL_FBC_A? Or is the parametrization just for
>> completeness?
>
> If it's not present in future hw then I guess it won't have a second
> instance. But I might prefer parametrizing all of them anyway.
Up to you. R-b stands as long as we don't break gvt build.
BR,
Jani.
>
>>
>> This one is only used in gvt, anyway. And that actually makes me wonder
>> if this should be breaking the build. Does CI not have gvt enabled?
>
> Hmm. I thought it was enabled in CI, but maybe not. I've often broken
> gvt with register define changes but I've always caught it before
> pushing. I think I have gvt enabled in my "make sure all commits build
> before I push" test config, so maybe that's where I caught most of them.
>
> Tomi, can we enable gvt in ci builds to make sure it at least still
> builds?
>
>> > #define DPFC_RECOMP_STALL_EN REG_BIT(27)
>> > #define DPFC_RECOMP_STALL_WM_MASK REG_GENMASK(26, 16)
>> > #define DPFC_RECOMP_TIMER_COUNT_MASK REG_GENMASK(5, 0)
>> > -#define DPFC_STATUS _MMIO(0x3210)
>> > -#define ILK_DPFC_STATUS _MMIO(0x43210)
>> > +#define DPFC_STATUS _MMIO(0x3210)
>> > +#define ILK_DPFC_STATUS(fbc_id) _MMIO_PIPE((fbc_id), 0x43210, 0x43250)
>>
>> Ditto, apart from the gvt part.
>
> Looks like this too might be gone in adl-p.
>
>>
>> > #define DPFC_INVAL_SEG_MASK REG_GENMASK(26, 16)
>> > #define DPFC_COMP_SEG_MASK REG_GENMASK(10, 0)
>> > -#define DPFC_STATUS2 _MMIO(0x3214)
>> > -#define ILK_DPFC_STATUS2 _MMIO(0x43214)
>> > +#define DPFC_STATUS2 _MMIO(0x3214)
>> > +#define ILK_DPFC_STATUS2(fbc_id) _MMIO_PIPE((fbc_id), 0x43214, 0x43254)
>> > #define DPFC_COMP_SEG_MASK_IVB REG_GENMASK(11, 0)
>> > -#define DPFC_FENCE_YOFF _MMIO(0x3218)
>> > -#define ILK_DPFC_FENCE_YOFF _MMIO(0x43218)
>> > -#define DPFC_CHICKEN _MMIO(0x3224)
>> > -#define ILK_DPFC_CHICKEN _MMIO(0x43224)
>> > +#define DPFC_FENCE_YOFF _MMIO(0x3218)
>> > +#define ILK_DPFC_FENCE_YOFF(fbc_id) _MMIO_PIPE((fbc_id), 0x43218, 0x43258)
>>
>> Ditto.
>
> I think this got nuked in ivb. It's not really used on snb either
> since snb introduced the system agent version of this register, but
> I'm pretty sure the old register was still present in the snb hw.
--
Jani Nikula, Intel Open Source Graphics Center
More information about the Intel-gfx
mailing list