- 25 May, 2021 8 commits
-
-
José Roberto de Souza authored
DC3CO is allowed in all the combinations between pipe and port A and B on alderlake-P. BSpec: 49196 Cc: Anshuman Gupta <anshuman.gupta@intel.com> Cc: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com> Cc: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by:
José Roberto de Souza <jose.souza@intel.com> Reviewed-by:
Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210524214805.259692-4-jose.souza@intel.com
-
José Roberto de Souza authored
Commit c457d9cf ("drm/i915: Make sure we have enough memory bandwidth on ICL") assumes that we always have a non-zero dram_info->channels and uses it as a divisor. We need num memory channels to be at least 1 for sane bw limits checking, even when PCode returns 0 or there is a error reading it, so lets force it to 1 in this case. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by:
Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by:
Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
-
José Roberto de Souza authored
On ADL-P TC cold is exited and blocked when legacy aux is powered, that is exacly the same of what ICL need for static TC ports. TODO: When a TBT hub or monitor is connected it will cause TBT and legacy aux to be powered at the same time, hopefully this will not cause any issues but if it do, some rework will be needed. v2: - skip icl_tc_port_assert_ref_held() warn on, adl-p uses aux to block TC cold v3: - Drop icl_tc_port_assert_ref_held() earlier return for adl_p, not needed anymore - Set timeout_expected when enabling aux power well as port could be disconnected when tc_cold_block() is called BSpec: 55480 Cc: Imre Deak <imre.deak@intel.com> Signed-off-by:
Matt Roper <matthew.d.roper@intel.com> Reviewed-by:
-
José Roberto de Souza authored
MODULAR_FIA_MASK is set in adl_p so we can drop this ealier return and read registers. Also to avoid warnings from icl_tc_port_assert_ref_held() when calling tc_cold_block() in this functions it is necessary to held the lock. Cc: Imre Deak <imre.deak@intel.com> Signed-off-by:
-
Imre Deak authored
Similarly to GGTT VMAs, DPT VMAs can be also a remapped or rotated view of the mapped object, so make sure we debug print the details for these views as well besides the normal view. While at it also fix the debug print for the VMA type of DPT VMAs. Signed-off-by:
Imre Deak <imre.deak@intel.com> Reviewed-by:
-
Imre Deak authored
An object mapped via DPT can have remapped and rotated VMA instances besides the normal VMA instance, similarly to GGTT VMA instances. Adjust the corresponding VMA lookup asserts. While at it also check if a DPT VM is passed incorrectly to i915_vm_to_ppgtt(). Signed-off-by:
-
Imre Deak authored
All DPT FB color plane surface base addresses must be 2MB aligned. On ADL_P this means that the offsets in CCS FB object must be also 2MB aligned. Adjusting unaligned offsets for these FBs during commit time (compensating with the x/y offsets) doesn't work, since the big alignment would most probably lead to an x/y offset mismatch error between the main and CCS planes. We can overcome this limitation by remapping CCS FBs, so that each color plane is at an aligned offset, leaving x/y for each plane unadjusted during commit and so not causing an x/y mismatch error. However remapping for CCS FBs will be done as a follow-up, so for now require that user space allocates the FB obj with properly aligned planes. v2: s/SZ_2M/512*4k/ for clarity. (Ville) Signed-off-by:
Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210524172703.2113058-1-imre.deak@intel.com
-
Lucas De Marchi authored
Fix table returned when port_clock > 270000: drivers/gpu/drm/i915/display/intel_ddi_buf_trans.c:752:47: error: variable 'adlp_dkl_phy_dp_ddi_trans_hbr2_hbr3' is not needed and will not be emitted [-Werror,-Wunneeded-internal-declaration] Initial version of the patch had it in a single table, but on second version the table got split, but we continued to reference just one of them. Fixes: ca962882 ("drm/i915/adl_p: Define and use ADL-P specific DP translation tables") Reported-by:
kernel test robot <lkp@intel.com> Signed-off-by:
Anusha Srivatsa <anusha.srivatsa@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210521005209.4058702-1-lucas.demarchi@intel.com
-
- 24 May, 2021 1 commit
-
-
Anusha Srivatsa authored
Propogate changes to macros name containing CSR_* to DMC_* from display side. Fixes: 0633cdcb ("drm/i915/dmc: Rename macro names containing csr") Cc: intel-gvt-dev@lists.freedesktop.org Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Reviewed-by:
Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210521174047.3861-1-anusha.srivatsa@intel.com
-
- 21 May, 2021 1 commit
-
-
Imre Deak authored
The driver currently disables the LTTPR non-transparent link training mode for sinks with a DPCD_REV<1.4, based on the following description of the LTTPR DPCD register range in DP standard 2.0 (at the 0xF0000 register description): "" LTTPR-related registers at DPCD Addresses F0000h through F02FFh are valid only for DPCD r1.4 (or higher). """ The transparent link training mode should still work fine, however the implementation for this in some retimer FWs seems to be broken, see the References: link below. After discussions with DP standard authors the above "DPCD r1.4" does not refer to the DPCD revision (stored in the DPCD_REV reg at 0x00000), rather to the "LTTPR field data structure revision" stored in the 0xF0000 reg. An update request has been filed at vesa.org (see wg/Link/documentComment/3746) for the upcoming v2.1 specification to clarify the above description along the following lines: """ LTTPR-related registers at DPCD Addresses F0000h through F02FFh are valid only for LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV 1.4 (or higher) """ Based on my tests Windows uses the non-transparent link training mode for DPCD_REV==1.2 sinks as well (so presumably for all DPCD_REVs), and forcing it to use transparent mode on ICL/TGL platforms leads to the same LT failure as reported at the References: link. Based on the above let's assume that the transparent link training mode is not well tested/supported and align the code to the correct interpretation of what the r1.4 version refers to. Reported-and-tested-by:
Casey Harkins <caseyharkins@gmail.com> Tested-by:
Khaled Almahallawy <khaled.almahallawy@intel.com> References: https://gitlab.freedesktop.org/drm/intel/-/issues/3415 Fixes: 264613b4 ("drm/i915: Disable LTTPR support when the DPCD rev < 1.4") Cc: <stable@vger.kernel.org> # v5.11+ Signed-off-by:
-
- 20 May, 2021 22 commits
-
-
Anusha Srivatsa authored
ADL_P has same memory characteristics as ADL_S platform. Bspec: 64631 Cc: José Roberto de Souza <jose.souza@intel.com> Cc: Clint Taylor <clinton.a.taylor@intel.com> Signed-off-by:
-
Anusha Srivatsa authored
When scalers are enabled, we need to program underrun bubble counter to 0x50 to avoid Soft Pipe A underruns. Make sure other bits dont get overwritten. Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Clint Taylor <clinton.a.taylor@intel.com> Cc: José Roberto de Souza <jose.souza@intel.com> Signed-off-by:
-
Imre Deak authored
On ADL_P besides programming the PLL accordingly the DP/HDMI link rate should be also programmed to the DDI_BUF_CTL register, do that. Cc: José Roberto de Souza <jose.souza@intel.com> Signed-off-by:
-
Anusha Srivatsa authored
The clocks in ALD_P is similar to that of TGL. The combo PLLs use the same DPLL0, DPLL1 and TBT_PLL. This patch adds the helper function intel_mg_pll_enable_reg() which is similar to intel_combo_pll_enable_reg() for being lookup place for PLL_ENABLE register in combo phy cases. Bspec: 55409,55316 Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Clinton Taylor <clinton.a.taylor@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: José Roberto de Souza <jose.souza@intel.com> Signed-off-by:
-
Mika Kahola authored
Define and use DP voltage swing and pre-emphasis translation tables for ADL-P. v2: - Update according to recent bspec updates; there are now separate tables for RBR/HBR and HBR2/HBR3. (Anusha) BSpec: 54956 Cc: Imre Deak <imre.deak@intel.com> Signed-off-by:
Mika Kahola <mika.kahola@intel.com> Signed-off-by:
-
Gwan-gyeong Mun authored
In order to reuse code of PSR interrupt error check on other PSR functions, it adds psr_interrupt_error_check() function. Cc: José Roberto de Souza <jose.souza@intel.com> Signed-off-by:
Gwan-gyeong Mun <gwan-gyeong.mun@intel.com> Signed-off-by:
-
Gwan-gyeong Mun authored
It replaces dc3co_enabled with dc3co_exitline on intel_psr struct. And it saves dc3co_exitline, not dc3co_enabled, so we can use dc3co_exitline without intel_crtc_state on other psr internal function like as intel_psr_enable_source(). Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: José Roberto de Souza <jose.souza@intel.com> Cc: Anshuman Gupta <anshuman.gupta@intel.com> Signed-off-by:
Anshuman Gupta <anshuman.gupta@intel.com> Signed-off-by:
-
Mika Kahola authored
Today when the DSI controller is paired with the Combo-PHY it uses the high-speed (HS) Word clock for its low power (LP) transmit PPI communication to the DPHY. The interface signaling only changes state at an Escape clock frequency (i.e. its effectively running on a virtual Tx Escape clock that is controlled by counters w/in the controller), but all the interface flops are running off the HS clock. This has the following drawbacks: * It is a deviation from the PPI spec which assumes signaling is running on a physical Escape clock * The PV timings are over constrained (HS timed to 312.5MHz vs. an Escape clock of 20MHz max) This feature is proposing to change the LP Tx communication between the controller and the DPHY from a virtual Tx Escape clock to a physical clock. To do this we need to program two "M" divisors. One for the usual DSI_ESC_CLK_DIV and DPHY_ESC_CLK_DIV register and one for MIPIO_DWORD8. For DSI_ESC_CLK_DIV and DPHY_ESC_CLK_DIV registers the "M" is calculated as following Nt = ceil(f_link/160) (theoretical word clock) Nact = max[3, Nt + (Nt + 1)%2] (actual word clock) M = Nact * 8 For MIPIO_DWORD8 register, the divisor "M" is calculated as following M = (Nact - 1)/2 BSpec: 55171 Cc: Vandita Kulkarni <vandita.kulkarni@intel.com> Signed-off-by:
Vandita Kulkarni <vandita.kulkarni@intel.com> Signed-off-by:
-
Vandita Kulkarni authored
Update MBUS_CTL register if the 2 mbus can be joined as per the current DDB allocation and active pipes, also update hashing mode and pipe select bits as per the sequence mentioned in the bspec. Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Cc: José Roberto de Souza <jose.souza@intel.com> Signed-off-by:
Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> Signed-off-by:
-
Ville Syrjälä authored
The dbuf slices are going to be split across several MBUS units. The actual dbuf programming will use offsets relative to the MBUS unit. To accommodate that we shall store the MBUS relative offsets into the dbuf_state->ddb[] and crtc_state->plane_ddb*[]. For crtc_state->wm.skl.ddb however we want to stick to global offsets as we use this to sanity check that the ddb allocations don't overlap between pipes. Cc: Clint Taylor <Clinton.A.Taylor@intel.com> Signed-off-by:
-
Vandita Kulkarni authored
On adlp the two mbuses have two display pipes and two DBUFS, Pipe A and D on Mbus1 and Pipe B and C on Mbus2. The Mbus can be joined and all the DBUFS can be used on Pipe A or B. Bspec: 49255 Cc: Anusha Srivatsa <anusha.srivatsa@intel.com> Signed-off-by:
-
José Roberto de Souza authored
Alderlake-P don't have programing sequences for MBUS or DBUF during display initializaiton, instead it requires programing to those registers during modeset because it to depend on the pipes left enabled. Bspec: 49213 Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by:
-
José Roberto de Souza authored
ADL-P have basically the same TC connection and disconnection sequences as ICL and TGL, the major difference is the new registers. So here adding functions without the icl prefix in the name and making the new functions call the platform specific function to access the correct register. v2: - Retain DDI TC PHY ownership flag during modesetting. BSpec: 55480 Cc: Imre Deak <imre.deak@intel.com> Signed-off-by:
-
Anusha Srivatsa authored
The SoC has 6 DDI ports(DDI A,DDI B and DDI TC1-4. The first two are connected to combo phys while the rest are connected to TC phys. Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Clinton Taylor <clinton.a.taylor@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Swathi Dhanavanthri <swathi.dhanavanthri@intel.com> Signed-off-by:
-
Matt Roper authored
XE_LPD reduces the number of regular watermark latency levels from 8 to 6 on non-dgfx platforms. However the hardware also adds a special purpose SAGV wateramrk (and an accompanying transition watermark) that will be used by the hardware in place of the level 0 values during SAGV transitions. Bspec: 49325, 49326, 50419 Cc: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by:
-
Vandita Kulkarni authored
Add the qp table for 444 formats, for 8bpc, 10bpc and 12bpc, as given by the VESA C model for DSC 1.1 v2: - Add include guard to header (Jani) - Move the big tables to a .c file (Chris, Jani, Lucas) v3: - Make tables 'static const' and add lookup functions to index into them. (Jani) v3.1: - Include missing .h file. Cc: Manasi Navare <manasi.d.navare@intel.com> Signed-off-by:
-
Vandita Kulkarni authored
Add methods to calculate rc parameters for all bpps, against the fixed arrays that we already have for 8,10,12 valid o/p bpps, to cover RGB 444 formats. Our hw doesn't support YUV compression yet. The calculations used here are from VESA C model for DSC 1.1 v2: - Checkpatch fixes Cc: Manasi Navare <manasi.d.navare@intel.com> Cc: Juha-Pekka Heikkil <juha-pekka.heikkila@intel.com> Signed-off-by:
Mohammed Khajapasha <mohammed.khajapasha@intel.com> Signed-off-by:
Manasi Navare <manasi.d.navare@intel.com> Signed-off-by:
-
Anusha Srivatsa authored
Finally, rename the header and source file from csr to dmc. v2: Add file rename in Documentation. - Place headers in orders. (Jani) Cc: Jani Nikula <jani.nikula@linux.intel.com> Signed-off-by:
-
Anusha Srivatsa authored
No functional change. Cc: Jani Nikula <jani.nikula@linux.intel.com> Signed-off-by:
-
Anusha Srivatsa authored
Rename all occurences of CSR_* with DMC_* Cc: Jani Nikula <jani.nikula@linux.intel.com> Signed-off-by:
-
Anusha Srivatsa authored
No functional change. Cc: Jani Nikula <jani.nikula@linux.intel.com> Signed-off-by:
-
Anusha Srivatsa authored
No functional change. v2: Chchpatch fixes. Cc: Jani Nikula <jani.nikula@linux.intel.com> Signed-off-by:
-
- 18 May, 2021 2 commits
-
-
José Roberto de Souza authored
If the do while loop breaks in 'if (!sg_dma_len(sgl))' in the first iteration, err is uninitialized causing a wrong call to zap_vma_ptes(). But that is impossible to happen as a scatterlist must have at least one valid segment. Anyways to avoid more reports from static checkers initializing ret here. Fixes: b12d691e ("i915: fix remap_io_sg to verify the pgprot") Reviewed-by:
Christoph Hellwig <hch@lst.de> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by:
James Ausmus <james.ausmus@intel.com> Signed-off-by:
-
Simon Rettberg authored
When resetting CACHE_MODE registers, don't enable HiZ Raw Stall Optimization on Ivybridge GT1 and Baytrail, as it causes severe glitches when rendering any kind of 3D accelerated content. This optimization is disabled on these platforms by default according to official documentation from 01.org. Fixes: ef99a60f ("drm/i915/gt: Clear CACHE_MODE prior to clearing residuals") BugLink: https://gitlab.freedesktop.org/drm/intel/-/issues/3081 BugLink: https://gitlab.freedesktop.org/drm/intel/-/issues/3404 BugLink: https://gitlab.freedesktop.org/drm/intel/-/issues/3071Reviewed-By:
Manuel Bentele <development@manuel-bentele.de> Signed-off-by:
Simon Rettberg <simon.rettberg@rz.uni-freiburg.de> Reviewed-by:
Dave Airlie <airlied@redhat.com> Signed-off-by:
Rodrigo Vivi <rodrigo.vivi@intel.com> [Rodrigo removed invalid Fixes line] Link: https://patchwork.freedesktop.org/patch/msgid/20210426161124.2b7fd708@dellnichtsogutkiste
-
- 17 May, 2021 1 commit
-
-
Rodrigo Vivi authored
Time to get back in sync... Signed-off-by:
-
- 15 May, 2021 5 commits
-
-
José Roberto de Souza authored
Buffer compression is not usable in A stepping. Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Anusha Srivatsa <anusha.srivatsa@intel.com> Cc: Clinton A Taylor <clinton.a.taylor@intel.com> Cc: Juha-Pekka Heikkilä <juha-pekka.heikkila@intel.com> Signed-off-by:
-
José Roberto de Souza authored
Implementation details are in the HSD 22011320316, requiring CD clock to be at least 307MHz to make DC states to work. Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Anusha Srivatsa <anusha.srivatsa@intel.com> Signed-off-by:
-
José Roberto de Souza authored
Adding a new hook to ADL-P just to avoid another platform check in gen12lp_init_clock_gating() but also open to it. BSpec: 54369 Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Anusha Srivatsa <anusha.srivatsa@intel.com> Signed-off-by:
-
José Roberto de Souza authored
This will allow us to better implement workarounds. Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by:
-
Animesh Manna authored
Respective bit for master or slave to be set for uncompressed bigjoiner in dss_ctl1 register. Cc: Manasi Navare <manasi.d.navare@intel.com> Signed-off-by:
Animesh Manna <animesh.manna@intel.com> Signed-off-by:
-
