Title: DC Patches March 18, 2022
cover Title: DC Patches March 18, 2022
This DC patchset brings improvements in multiple areas. In summary, we highlight:
* HDCP SEND AKI INIT error
* fix audio format not updated after edid updated
* Reduce stack size
* FEC check in timing validation
* Add fSMC_MSG_SetDtbClk support
* Update VTEM Infopacket definition
* [FW Promotion] Release 0.0.109.0
* Add support for zstate during extended vblank
* remove destructive verify link for TMDS
* move FPU related code from dcn31 to dml/dcn31 folder
* move FPU related code from dcn315 to dml/dcn31 folder
* move FPU related code from dcn316 to dml/dcn31 folder
Ahmad Othman (1):
drm/amd/display: HDCP SEND AKI INIT error
Anthony Koo (1):
drm/amd/display: [FW Promotion] Release 0.0.109.0
Aric Cyr (1):
drm/amd/display: 3.2.178
Charlene Liu (2):
drm/amd/display: fix audio format not updated after edid updated drm/amd/display: remove destructive verify link for TMDS
Chiawen Huang (1):
drm/amd/display: FEC check in timing validation
Gabe Teeger (1):
drm/amd/display: Add support for zstate during extended vblank
Leo (Hanghong) Ma (1):
drm/amd/display: Update VTEM Infopacket definition
Melissa Wen (3):
drm/amd/dicplay: move FPU related code from dcn31 to dml/dcn31 folder drm/amd/display: move FPU related code from dcn315 to dml/dcn31 folder drm/amd/display: move FPU related code from dcn316 to dml/dcn31 folder
Oliver Logush (1):
drm/amd/display: Add fSMC_MSG_SetDtbClk support
Rodrigo Siqueira (1):
drm/amd/display: Reduce stack size
.../display/dc/clk_mgr/dcn315/dcn315_smu.c | 19 +- .../display/dc/clk_mgr/dcn315/dcn315_smu.h | 4 +- drivers/gpu/drm/amd/display/dc/core/dc.c | 23 + drivers/gpu/drm/amd/display/dc/core/dc_link.c | 3 +- .../gpu/drm/amd/display/dc/core/dc_resource.c | 4 +- drivers/gpu/drm/amd/display/dc/dc.h | 8 +- drivers/gpu/drm/amd/display/dc/dc_stream.h | 2 + .../amd/display/dc/dcn10/dcn10_hw_sequencer.c | 8 +- .../drm/amd/display/dc/dcn20/dcn20_hwseq.c | 12 + drivers/gpu/drm/amd/display/dc/dcn31/Makefile | 26 - .../gpu/drm/amd/display/dc/dcn31/dcn31_hubp.c | 8 + .../drm/amd/display/dc/dcn31/dcn31_resource.c | 356 +------- .../drm/amd/display/dc/dcn31/dcn31_resource.h | 4 +- .../gpu/drm/amd/display/dc/dcn315/Makefile | 26 - .../amd/display/dc/dcn315/dcn315_resource.c | 232 +---- .../amd/display/dc/dcn315/dcn315_resource.h | 3 + .../gpu/drm/amd/display/dc/dcn316/Makefile | 26 - .../amd/display/dc/dcn316/dcn316_resource.c | 231 +---- .../amd/display/dc/dcn316/dcn316_resource.h | 3 + drivers/gpu/drm/amd/display/dc/dml/Makefile | 2 + .../drm/amd/display/dc/dml/dcn20/dcn20_fpu.c | 18 +- .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 863 ++++++++++++++++++ .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 44 + .../dc/dml/dcn31/display_rq_dlg_calc_31.c | 13 + .../amd/display/dc/dml/display_mode_structs.h | 2 + drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h | 3 + .../gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 4 +- .../gpu/drm/amd/display/modules/hdcp/hdcp.c | 38 +- .../gpu/drm/amd/display/modules/hdcp/hdcp.h | 8 + .../drm/amd/display/modules/inc/mod_hdcp.h | 2 +- .../display/modules/info_packet/info_packet.c | 5 +- 31 files changed, 1085 insertions(+), 915 deletions(-) create mode 100644 drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c create mode 100644 drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h
Signature · 1 line
2.35.1
PATCH 1/13 drm/amd/display: HDCP SEND AKI INIT error
From: Ahmad Othman <[email protected]>
[why]
HDCP sends AKI INIT error in case of multiple display on dock
[how]
Added new checks and method to handfle display adjustment for multiple display cases
--- .../gpu/drm/amd/display/modules/hdcp/hdcp.c | 38 ++++++++++++++++++- .../gpu/drm/amd/display/modules/hdcp/hdcp.h | 8 ++++ .../drm/amd/display/modules/inc/mod_hdcp.h | 2 +- 3 files changed, 46 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c +37 -1
@@ -251,6 +251,33 @@ static enum mod_hdcp_status reset_connection(struct mod_hdcp *hdcp,return status; } +static enum mod_hdcp_status update_display_adjustments (struct mod_hdcp *hdcp, +struct mod_hdcp_display *display, +struct mod_hdcp_display_adjustment *adj) +{ +enum mod_hdcp_status status =MOD_HDCP_STATUS_NOT_IMPLEMENTED ; + +if (is_in_authenticated_states (hdcp) && +is_dp_mst_hdcp (hdcp) && + display->adjust.disable == true && + adj->disable == false) { + display->adjust.disable = false; +if (is_hdcp1 (hdcp)) + status =mod_hdcp_hdcp1_enable_dp_stream_encryption (hdcp); +else if (is_hdcp2 (hdcp)) + status =mod_hdcp_hdcp2_enable_dp_stream_encryption (hdcp); + +if (status !=MOD_HDCP_STATUS_SUCCESS ) + display->adjust.disable = true; + } + +if (status ==MOD_HDCP_STATUS_SUCCESS && +memcmp (adj, &display->adjust, +sizeof (struct mod_hdcp_display_adjustment )) !=0 ) + status =MOD_HDCP_STATUS_NOT_IMPLEMENTED ; + +return status; +}/* * Implementation of functions in mod_hdcp.h */
@@ -391,7 +418,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp,return status; } -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust,
@@ -419,6 +446,15 @@ enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, goto out; } +if (memcmp (link_adjust, &hdcp->connection.link.adjust, +sizeof (struct mod_hdcp_link_adjustment )) ==0 && +memcmp (display_adjust, &display->adjust, +sizeof (struct mod_hdcp_display_adjustment )) !=0 ) { + status =update_display_adjustments (hdcp, display, display_adjust); +if (status !=MOD_HDCP_STATUS_NOT_IMPLEMENTED ) + goto out; + } +/* stop current authentication */ status =reset_authentication (hdcp, output);if (status !=MOD_HDCP_STATUS_SUCCESS )
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +8 -0
@@ -445,6 +445,14 @@ static inline uint8_t is_in_hdcp2_dp_states(struct mod_hdcp *hdcp) current_state(hdcp) <= HDCP2_DP_STATE_END); } +static inline uint8_t is_in_authenticated_states (struct mod_hdcp *hdcp) +{ +return (current_state (hdcp) ==D1_A4_AUTHENTICATED || +current_state (hdcp) ==H1_A45_AUTHENTICATED || +current_state (hdcp) ==D2_A5_AUTHENTICATED || +current_state (hdcp) ==H2_A5_AUTHENTICATED ); +} +static inline uint8_t is_hdcp1 (struct mod_hdcp *hdcp) {return (is_in_hdcp1_states (hdcp) ||is_in_hdcp1_dp_states (hdcp));
drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +1 -1
@@ -294,7 +294,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_output *output);/* called per display to apply new authentication adjustment */ -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust,
Signature · 1 line
2.35.1
Re: [PATCH 01/13] drm/amd/display: HDCP SEND AKI INIT error
Dear Alex, dear Ahmad,
Thank you for the patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 1 line
From: Ahmad Othman <[email protected]>
Could you please make the commit message summary/title a statement by adding a verb (imperative mood) [1]. Maybe:
drm/amd/display: Fix HDCP SEND AKI INIT error
Quoted 2 lines
[why]
HDCP sends AKI INIT error in case of multiple display on dock
What is the test setup exactly, and how can the error be reproduced? Does Linux log something?
Quoted 2 lines
[how]
Added new checks and method to handfle display adjustment
s/Added/Add/
s/handfle/handle/
Quoted 1 line
for multiple display cases
Why are these checks and methods correct, and what do they try to achieve? Is it the HDCP(?) specification?
Quoted 3 lines
Reviewed-by Wenjing Liu <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Ahmad Othman <[email protected]>
Could the order be reversed, so it’s clear that the Signed-off-by line came first and not after the review? Or is it actually signed off after the review again?
Quoted 75 lines
--- .../gpu/drm/amd/display/modules/hdcp/hdcp.c | 38 ++++++++++++++++++- .../gpu/drm/amd/display/modules/hdcp/hdcp.h | 8 ++++ .../drm/amd/display/modules/inc/mod_hdcp.h | 2 +- 3 files changed, 46 insertions(+), 2 deletions(-)drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c +37
-1@@ -251,6 +251,33 @@ static enum mod_hdcp_status reset_connection(struct mod_hdcp *hdcp,return status; } +static enum mod_hdcp_status update_display_adjustments (struct mod_hdcp *hdcp, +struct mod_hdcp_display *display, +struct mod_hdcp_display_adjustment *adj) +{ +enum mod_hdcp_status status =MOD_HDCP_STATUS_NOT_IMPLEMENTED ; + +if (is_in_authenticated_states (hdcp) && +is_dp_mst_hdcp (hdcp) && + display->adjust.disable == true && + adj->disable == false) { + display->adjust.disable = false; +if (is_hdcp1 (hdcp)) + status =mod_hdcp_hdcp1_enable_dp_stream_encryption (hdcp); +else if (is_hdcp2 (hdcp)) + status =mod_hdcp_hdcp2_enable_dp_stream_encryption (hdcp); + +if (status !=MOD_HDCP_STATUS_SUCCESS ) + display->adjust.disable = true; + } + +if (status ==MOD_HDCP_STATUS_SUCCESS && +memcmp (adj, &display->adjust, +sizeof (struct mod_hdcp_display_adjustment )) !=0 ) + status =MOD_HDCP_STATUS_NOT_IMPLEMENTED ; + +return status; +}/* * Implementation of functions in mod_hdcp.h */ @@ -391,7 +418,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp,return status; } -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust,@@ -419,6 +446,15 @@ enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, goto out; } +if (memcmp (link_adjust, &hdcp->connection.link.adjust, +sizeof (struct mod_hdcp_link_adjustment )) ==0 && +memcmp (display_adjust, &display->adjust, +sizeof (struct mod_hdcp_display_adjustment )) !=0 ) { + status =update_display_adjustments (hdcp, display, display_adjust); +if (status !=MOD_HDCP_STATUS_NOT_IMPLEMENTED ) + goto out; + } +/* stop current authentication */ status =reset_authentication (hdcp, output);if (status !=MOD_HDCP_STATUS_SUCCESS )drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +0
-0@@ -445,6 +445,14 @@ static inline uint8_t is_in_hdcp2_dp_states(struct mod_hdcp *hdcp) current_state(hdcp) <= HDCP2_DP_STATE_END);
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h static inline uint8_t is_in_hdcp2_dp_states(struct mod_hdcp *hdcp) · line 455
Quoted 10 lines
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +8
-0} +static inline uint8_t is_in_authenticated_states (struct mod_hdcp *hdcp) +{ +return (current_state (hdcp) ==D1_A4_AUTHENTICATED || +current_state (hdcp) ==H1_A45_AUTHENTICATED || +current_state (hdcp) ==D2_A5_AUTHENTICATED || +current_state (hdcp) ==H2_A5_AUTHENTICATED ); +} +
The compiler is probably going to optimize the four current_state calls away, but maybe use a helper variable, so it’s clear for the reader the same function is each time. Also, why not put the helper into drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h?
Quoted 8 lines
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +0
-0static inline uint8_t is_hdcp1 (struct mod_hdcp *hdcp) {return (is_in_hdcp1_states (hdcp) ||is_in_hdcp1_dp_states (hdcp));drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +0
-0
drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, · line 300
Quoted 8 lines
drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +1
-1@@ -294,7 +294,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_output *output);/* called per display to apply new authentication adjustment */ -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust,
Kind regards,
Paul
Re: [PATCH 01/13] drm/amd/display: HDCP SEND AKI INIT error
[AMD Official Use Only]
Hi Paul,
Thanks for your feedbacks. I fixed many errors and typos you highlighted in this series. In cases where modification requires re-testing we or anyone can have follow-up patches in the future.
________________________________
From: Paul Menzel <[email protected]>
Sent: 19 March 2022 01:16
To: Hung, Alex <[email protected]>; Othman, Ahmad <[email protected]>
Cc: [email protected] <[email protected]>; Wang, Chao-kai (Stylon) <[email protected]>; Li, Sun peng (Leo) <[email protected]>; Wentland, Harry <[email protected]>; Zhuo, Qingqing (Lillian) <[email protected]>; Siqueira, Rodrigo <[email protected]>; Li, Roman <[email protected]>; Chiu, Solomon <[email protected]>; Pillai, Aurabindo <[email protected]>; Lin, Wayne <[email protected]>; Liu, Wenjing <[email protected]>; Lakha, Bhawanpreet <[email protected]>; Gutierrez, Agustin <[email protected]>; Kotarac, Pavle <[email protected]>
Subject: Re: [PATCH 01/13] drm/amd/display: HDCP SEND AKI INIT error
Dear Alex, dear Ahmad,
Thank you for the patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 1 line
From: Ahmad Othman <[email protected]>
Could you please make the commit message summary/title a statement by adding a verb (imperative mood) [1]. Maybe:
drm/amd/display: Fix HDCP SEND AKI INIT error
Quoted 2 lines
[why]
HDCP sends AKI INIT error in case of multiple display on dock
What is the test setup exactly, and how can the error be reproduced? Does Linux log something?
Quoted 2 lines
[how]
Added new checks and method to handfle display adjustment
s/Added/Add/
s/handfle/handle/
Quoted 1 line
for multiple display cases
Why are these checks and methods correct, and what do they try to achieve? Is it the HDCP(?) specification?
Quoted 3 lines
Reviewed-by Wenjing Liu <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Ahmad Othman <[email protected]>
Could the order be reversed, so it’s clear that the Signed-off-by line came first and not after the review? Or is it actually signed off after the review again?
Quoted 75 lines
--- .../gpu/drm/amd/display/modules/hdcp/hdcp.c | 38 ++++++++++++++++++- .../gpu/drm/amd/display/modules/hdcp/hdcp.h | 8 ++++ .../drm/amd/display/modules/inc/mod_hdcp.h | 2 +- 3 files changed, 46 insertions(+), 2 deletions(-)drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c +37
-1@@ -251,6 +251,33 @@ static enum mod_hdcp_status reset_connection(struct mod_hdcp *hdcp,return status; } +static enum mod_hdcp_status update_display_adjustments (struct mod_hdcp *hdcp, +struct mod_hdcp_display *display, +struct mod_hdcp_display_adjustment *adj) +{ +enum mod_hdcp_status status =MOD_HDCP_STATUS_NOT_IMPLEMENTED ; + +if (is_in_authenticated_states (hdcp) && +is_dp_mst_hdcp (hdcp) && + display->adjust.disable == true && + adj->disable == false) { + display->adjust.disable = false; +if (is_hdcp1 (hdcp)) + status =mod_hdcp_hdcp1_enable_dp_stream_encryption (hdcp); +else if (is_hdcp2 (hdcp)) + status =mod_hdcp_hdcp2_enable_dp_stream_encryption (hdcp); + +if (status !=MOD_HDCP_STATUS_SUCCESS ) + display->adjust.disable = true; + } + +if (status ==MOD_HDCP_STATUS_SUCCESS && +memcmp (adj, &display->adjust, +sizeof (struct mod_hdcp_display_adjustment )) !=0 ) + status =MOD_HDCP_STATUS_NOT_IMPLEMENTED ; + +return status; +}/* * Implementation of functions in mod_hdcp.h */ @@ -391,7 +418,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp,return status; } -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust,@@ -419,6 +446,15 @@ enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, goto out; } +if (memcmp (link_adjust, &hdcp->connection.link.adjust, +sizeof (struct mod_hdcp_link_adjustment )) ==0 && +memcmp (display_adjust, &display->adjust, +sizeof (struct mod_hdcp_display_adjustment )) !=0 ) { + status =update_display_adjustments (hdcp, display, display_adjust); +if (status !=MOD_HDCP_STATUS_NOT_IMPLEMENTED ) + goto out; + } +/* stop current authentication */ status =reset_authentication (hdcp, output);if (status !=MOD_HDCP_STATUS_SUCCESS )drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +0
-0@@ -445,6 +445,14 @@ static inline uint8_t is_in_hdcp2_dp_states(struct mod_hdcp *hdcp) current_state(hdcp) <= HDCP2_DP_STATE_END);
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h static inline uint8_t is_in_hdcp2_dp_states(struct mod_hdcp *hdcp) · line 455
Quoted 10 lines
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +8
-0} +static inline uint8_t is_in_authenticated_states (struct mod_hdcp *hdcp) +{ +return (current_state (hdcp) ==D1_A4_AUTHENTICATED || +current_state (hdcp) ==H1_A45_AUTHENTICATED || +current_state (hdcp) ==D2_A5_AUTHENTICATED || +current_state (hdcp) ==H2_A5_AUTHENTICATED ); +} +
The compiler is probably going to optimize the four current_state calls away, but maybe use a helper variable, so it’s clear for the reader the same function is each time. Also, why not put the helper into drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h?
Quoted 8 lines
drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +0
-0static inline uint8_t is_hdcp1 (struct mod_hdcp *hdcp) {return (is_in_hdcp1_states (hdcp) ||is_in_hdcp1_dp_states (hdcp));drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +0
-0
drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, · line 300
Quoted 8 lines
drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +1
-1@@ -294,7 +294,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_output *output);/* called per display to apply new authentication adjustment */ -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust,
Kind regards,
Paul
Re: [PATCH 01/13] drm/amd/display: HDCP SEND AKI INIT error
Dear Alex,
Am 25.03.22 um 21:43 schrieb Hung, Alex:
Quoted 1 line
Thanks for your feedback.
Thank you for your reply. For the future, it’d be great if you could use interleaved style [1], when replying, or when it’s a new topic, do not cite the previous message.
Quoted 3 lines
I fixed many errors and typos you highlighted in this series. In cases where modification requires re-testing we or anyone can have follow-up patches in the future.
Thank you. Did you resubmit them in *[PATCH 00/16] DC Patches March 25, 2022*? Unfortunately, they are not tagged as v2 and the emails do not contain an information what was changed between v1 and v2. Can you please send them again with the correct tags, and information included?
Kind regards,
Paul
PATCH 2/13 drm/amd/display: fix audio format not updated after edid updated
In reply to Alex HungFrom: Charlene Liu <[email protected]>
[why]
for the case edid change only changed audio format. driver still need to update stream.
--- drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc_resource.c +2 -2
@@ -1685,8 +1685,8 @@ bool dc_is_stream_unchanged(if (old_stream->ignore_msa_timing_param != stream->ignore_msa_timing_param)return false; -// Only Have Audio left to check whether it is same or not. This is a corner case for Tiled sinks -if (old_stream->audio_info.mode_count != stream->audio_info.mode_count) +/*compare audio info*/ +if (memcmp (&old_stream->audio_info, &stream->audio_info,sizeof (stream->audio_info)) !=0 )return false;return true;
Signature · 1 line
2.35.1
PATCH 3/13 drm/amd/display: Reduce stack size
In reply to Alex HungFrom: Rodrigo Siqueira <[email protected]>
Linux kernel enabled more compilation restrictions related to the stack size, which caused compilation failures in our code. This commit reduces the allocation size by allocating the required memory dynamically.
--- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +7 -1
@@ -2056,7 +2056,7 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size, {struct dc_context *dc_ctx = dc->ctx;int i, master =-1 , embedded =-1 ; -struct dc_crtc_timing hw_crtc_timing[MAX_PIPES ] = {0 }; +struct dc_crtc_timing *hw_crtc_timing;uint64_t phase[MAX_PIPES ];uint64_t modulo[MAX_PIPES ];unsigned int pclk;
@@ -2067,6 +2067,10 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size,uint32_t dp_ref_clk_100hz = dc->res_pool->dp_clock_source->ctx->dc->clk_mgr->dprefclk_khz*10 ; + hw_crtc_timing =kcalloc (MAX_PIPES ,sizeof (*hw_crtc_timing),GFP_KERNEL ); +if (!hw_crtc_timing) +return master; +if (dc->config.vblank_alignment_dto_params && dc->res_pool->dp_clock_source->funcs->override_dp_pix_clk) { embedded_h_total =
@@ -2130,6 +2134,8 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size, } } + +kfree (hw_crtc_timing);return master; }
Signature · 1 line
2.35.1
Re: [PATCH 03/13] drm/amd/display: Reduce stack size
Dear Alex, dear Rodrigo,
Thank you for the patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 4 lines
From: Rodrigo Siqueira <[email protected]>
Linux kernel enabled more compilation restrictions related to the stack size,
Could you please reference the commit, where that was introduced.
Quoted 1 line
which caused compilation failures in our code.
Please give one example failure message.
Quoted 2 lines
This commit reduces the allocation size by allocating the required memory dynamically.
Did you do measurements how much the stack size was reduced in the end? If so, it’d be great if you documented that in the commit message.
Quoted 33 lines
Reviewed-by Harry Wentland <[email protected]>Reviewed-by Aric Cyr <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Rodrigo Siqueira <[email protected]>--- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +5
-1@@ -2056,7 +2056,7 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size, {struct dc_context *dc_ctx = dc->ctx;int i, master =-1 , embedded =-1 ; -struct dc_crtc_timing hw_crtc_timing[MAX_PIPES ] = {0 }; +struct dc_crtc_timing *hw_crtc_timing;uint64_t phase[MAX_PIPES ];uint64_t modulo[MAX_PIPES ];unsigned int pclk;@@ -2067,6 +2067,10 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size,uint32_t dp_ref_clk_100hz = dc->res_pool->dp_clock_source->ctx->dc->clk_mgr->dprefclk_khz*10 ; + hw_crtc_timing =kcalloc (MAX_PIPES ,sizeof (*hw_crtc_timing),GFP_KERNEL ); +if (!hw_crtc_timing) +return master; +if (dc->config.vblank_alignment_dto_params && dc->res_pool->dp_clock_source->funcs->override_dp_pix_clk) { embedded_h_total =
drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c static int dcn10_align_pixel_clocks(struct dc *dc, int group_size, · line 2141
Quoted 8 lines
drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +2
-0@@ -2130,6 +2134,8 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size, } } + +kfree (hw_crtc_timing);return master; }
Kind regards,
Paul
PATCH 4/13 drm/amd/display: FEC check in timing validation
In reply to Alex HungFrom: Chiawen Huang <[email protected]>
[Why]
disable/enable leads fec mismatch between hw/sw fec state.
[How]
check fec status to fastboot on/off.
--- drivers/gpu/drm/amd/display/dc/core/dc.c | 4 ++++ 1 file changed, 4 insertions(+)
drivers/gpu/drm/amd/display/dc/core/dc.c +4 -0
@@ -1496,6 +1496,10 @@ bool dc_validate_boot_timing(const struct dc *dc,if (!link->link_enc->funcs->is_dig_enabled (link->link_enc))return false; +/* Check for FEC status*/ +if (link->link_enc->funcs->fec_is_active (link->link_enc)) +return false; + enc_inst = link->link_enc->funcs->get_dig_frontend (link->link_enc);if (enc_inst ==ENGINE_ID_UNKNOWN )
Signature · 1 line
2.35.1
Re: [PATCH 04/13] drm/amd/display: FEC check in timing validation
Dear Alex, dear Chiawen,
Thank you for your patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 4 lines
From: Chiawen Huang <[email protected]>
[Why]
disable/enable leads fec mismatch between hw/sw fec state.
1. Disable/enable of what? 2. How can this be reproduced? 3. s/fec/FEC/
Quoted 2 lines
[How]
check fec status to fastboot on/off.
What do you mean by “to fastboot on/off”?
Quoted 12 lines
Reviewed-by Anthony Koo <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Chiawen Huang <[email protected]>--- drivers/gpu/drm/amd/display/dc/core/dc.c | 4 ++++ 1 file changed, 4 insertions(+)drivers/gpu/drm/amd/display/dc/core/dc.c +0
-0
drivers/gpu/drm/amd/display/dc/core/dc.c bool dc_validate_boot_timing(const struct dc *dc, · line 1499
Quoted 4 lines
drivers/gpu/drm/amd/display/dc/core/dc.c +1
-0@@ -1496,6 +1496,10 @@ bool dc_validate_boot_timing(const struct dc *dc,if (!link->link_enc->funcs->is_dig_enabled (link->link_enc))return false; +/* Check for FEC status*/
Missing space before */.
drivers/gpu/drm/amd/display/dc/core/dc.c bool dc_validate_boot_timing(const struct dc *dc, · line 1505
Quoted 6 lines
drivers/gpu/drm/amd/display/dc/core/dc.c +3
-0+if (link->link_enc->funcs->fec_is_active (link->link_enc)) +return false; + enc_inst = link->link_enc->funcs->get_dig_frontend (link->link_enc);if (enc_inst ==ENGINE_ID_UNKNOWN )
Kind regards,
Paul
Re: [PATCH 04/13] drm/amd/display: FEC check in timing validation
Dear Alex,
Am 19.03.22 um 08:43 schrieb Paul Menzel:
Quoted 27 lines
Dear Alex, dear Chiawen,
Thank you for your patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 4 lines
From: Chiawen Huang <[email protected]>
[Why]
disable/enable leads fec mismatch between hw/sw fec state.1. Disable/enable of what?
2. How can this be reproduced?
3. s/fec/FEC/Quoted 2 lines
[How]
check fec status to fastboot on/off.What do you mean by “to fastboot on/off”?
Quoted 7 lines
Reviewed-by Anthony Koo <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Chiawen Huang <[email protected]>---
drivers/gpu/drm/amd/display/dc/core/dc.c| 4 ++++1 file changed, 4 insertions(+)
Quoted 16 lines
drivers/gpu/drm/amd/display/dc/core/dc.c bool dc_validate_boot_timing(const struct dc *dc, · line 1498
Quoted 8 lines
drivers/gpu/drm/amd/display/dc/core/dc.c +1
-0@@ -1496,6 +1496,10 @@ bool dc_validate_boot_timing(const struct dc *dc, if (!link->link_enc->funcs->is_dig_enabled(link->link_enc)) return false; + /* Check for FEC status*/Missing space before
*/.Quoted 5 lines
drivers/gpu/drm/amd/display/dc/core/dc.c +3
-0+ if (link->link_enc->funcs->fec_is_active(link->link_enc)) + return false; + enc_inst = link->link_enc->funcs->get_dig_frontend(link->link_enc); if (enc_inst == ENGINE_ID_UNKNOWN)
I just saw, that this patch was committed on March 25th, 2022 (commit 7d56a154e22f) with my comments ignored. Could you please explain why?
Kind regards,
Paul
PATCH 5/13 drm/amd/display: Add fSMC_MSG_SetDtbClk support
In reply to Alex HungFrom: Oliver Logush <[email protected]>
[why]
Needed to support dcn315
--- .../display/dc/clk_mgr/dcn315/dcn315_smu.c | 19 +++++++++++++++---- .../display/dc/clk_mgr/dcn315/dcn315_smu.h | 4 +++- 2 files changed, 18 insertions(+), 5 deletions(-)
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c +15 -4
@@ -80,8 +80,8 @@ static const struct IP_BASE NBIO_BASE = { { { { 0x00000000, 0x00000014, 0x00000D#define VBIOSSMC_MSG_SetDppclkFreq 0x06 ///< Set DPP clock frequency in MHZ#define VBIOSSMC_MSG_SetHardMinDcfclkByFreq 0x07 ///< Set DCF clock frequency hard min in MHZ#define VBIOSSMC_MSG_SetMinDeepSleepDcfclk 0x08 ///< Set DCF clock minimum frequency in deep sleep in MHZ -#define VBIOSSMC_MSG_SetPhyclkVoltageByFreq 0x09 ///< Set display phy clock frequency in MHZ in case VMIN does not support phy frequency -#define VBIOSSMC_MSG_GetFclkFrequency 0x0A ///< Get FCLK frequency, return frequemcy in MHZ +#define VBIOSSMC_MSG_GetDtbclkFreq 0x09 ///< Get display dtb clock frequency in MHZ in case VMIN does not support phy frequency +#define VBIOSSMC_MSG_SetDtbClk 0x0A ///< Set dtb clock frequency, return frequemcy in MHZ#define VBIOSSMC_MSG_SetDisplayCount 0x0B ///< Inform PMFW of number of display connected#define VBIOSSMC_MSG_EnableTmdp48MHzRefclkPwrDown 0x0C ///< To ask PMFW turn off TMDP 48MHz refclk during display off to save power#define VBIOSSMC_MSG_UpdatePmeRestore 0x0D ///< To ask PMFW to write into Azalia for PME wake up event
@@ -324,15 +324,26 @@ int dcn315_smu_get_dpref_clk(struct clk_mgr_internal *clk_mgr)return (dprefclk_get_mhz *1000 ); } -int dcn315_smu_get_smu_fclk (struct clk_mgr_internal *clk_mgr) +int dcn315_smu_get_dtbclk (struct clk_mgr_internal *clk_mgr) {int fclk_get_mhz =-1 ;if (clk_mgr->smu_present) { fclk_get_mhz =dcn315_smu_send_msg_with_param ( clk_mgr, - VBIOSSMC_MSG_GetFclkFrequency, + VBIOSSMC_MSG_GetDtbclkFreq,0 ); }return (fclk_get_mhz *1000 ); } + +void dcn315_smu_set_dtbclk (struct clk_mgr_internal *clk_mgr,bool enable) +{ +if (!clk_mgr->smu_present) +return ; + +dcn315_smu_send_msg_with_param ( + clk_mgr, + VBIOSSMC_MSG_SetDtbClk, + enable); +}
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.h +3 -1
@@ -37,6 +37,7 @@#define NUM_SOC_VOLTAGE_LEVELS 4#define NUM_DF_PSTATE_LEVELS 4 + typedef struct { uint16_t MinClock; // This is either DCFCLK or SOCCLK (in MHz) uint16_t MaxClock; // This is either DCFCLK or SOCCLK (in MHz)
@@ -124,5 +125,6 @@ void dcn315_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr);void dcn315_smu_request_voltage_via_phyclk (struct clk_mgr_internal *clk_mgr,int requested_phyclk_khz);void dcn315_smu_enable_pme_wa (struct clk_mgr_internal *clk_mgr);int dcn315_smu_get_dpref_clk (struct clk_mgr_internal *clk_mgr); -int dcn315_smu_get_smu_fclk (struct clk_mgr_internal *clk_mgr); +int dcn315_smu_get_dtbclk (struct clk_mgr_internal *clk_mgr); +void dcn315_smu_set_dtbclk (struct clk_mgr_internal *clk_mgr,bool enable);#endif /* DAL_DC_315_SMU_H_ */
Signature · 1 line
2.35.1
Re: [PATCH 05/13] drm/amd/display: Add fSMC_MSG_SetDtbClk support
Dear Alex, dear Oliver,
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 4 lines
From: Oliver Logush <[email protected]>
[why]
Needed to support dcn315
Please elaborate. What is wrong with dcn315_smu_get_smu_fclk()?
Quoted 57 lines
Reviewed-by Charlene Liu <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Oliver Logush <[email protected]>--- .../display/dc/clk_mgr/dcn315/dcn315_smu.c | 19 +++++++++++++++---- .../display/dc/clk_mgr/dcn315/dcn315_smu.h | 4 +++- 2 files changed, 18 insertions(+), 5 deletions(-)drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c +15
-4@@ -80,8 +80,8 @@ static const struct IP_BASE NBIO_BASE = { { { { 0x00000000, 0x00000014, 0x00000D#define VBIOSSMC_MSG_SetDppclkFreq 0x06 ///< Set DPP clock frequency in MHZ#define VBIOSSMC_MSG_SetHardMinDcfclkByFreq 0x07 ///< Set DCF clock frequency hard min in MHZ#define VBIOSSMC_MSG_SetMinDeepSleepDcfclk 0x08 ///< Set DCF clock minimum frequency in deep sleep in MHZ -#define VBIOSSMC_MSG_SetPhyclkVoltageByFreq 0x09 ///< Set display phy clock frequency in MHZ in case VMIN does not support phy frequency -#define VBIOSSMC_MSG_GetFclkFrequency 0x0A ///< Get FCLK frequency, return frequemcy in MHZ +#define VBIOSSMC_MSG_GetDtbclkFreq 0x09 ///< Get display dtb clock frequency in MHZ in case VMIN does not support phy frequency +#define VBIOSSMC_MSG_SetDtbClk 0x0A ///< Set dtb clock frequency, return frequemcy in MHZ#define VBIOSSMC_MSG_SetDisplayCount 0x0B ///< Inform PMFW of number of display connected#define VBIOSSMC_MSG_EnableTmdp48MHzRefclkPwrDown 0x0C ///< To ask PMFW turn off TMDP 48MHz refclk during display off to save power#define VBIOSSMC_MSG_UpdatePmeRestore 0x0D ///< To ask PMFW to write into Azalia for PME wake up event@@ -324,15 +324,26 @@ int dcn315_smu_get_dpref_clk(struct clk_mgr_internal *clk_mgr)return (dprefclk_get_mhz *1000 ); } -int dcn315_smu_get_smu_fclk (struct clk_mgr_internal *clk_mgr) +int dcn315_smu_get_dtbclk (struct clk_mgr_internal *clk_mgr) {int fclk_get_mhz =-1 ;if (clk_mgr->smu_present) { fclk_get_mhz =dcn315_smu_send_msg_with_param ( clk_mgr, - VBIOSSMC_MSG_GetFclkFrequency, + VBIOSSMC_MSG_GetDtbclkFreq,0 ); }return (fclk_get_mhz *1000 ); } + +void dcn315_smu_set_dtbclk (struct clk_mgr_internal *clk_mgr,bool enable) +{ +if (!clk_mgr->smu_present) +return ; + +dcn315_smu_send_msg_with_param ( + clk_mgr, + VBIOSSMC_MSG_SetDtbClk, + enable); +}drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.h +0
-0
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.h line 40
Quoted 4 lines
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.h +1
-0@@ -37,6 +37,7 @@#define NUM_SOC_VOLTAGE_LEVELS 4#define NUM_DF_PSTATE_LEVELS 4 +
Unrelated.
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.h void dcn315_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr); · line 130
Quoted 11 lines
drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.h +2
-1typedef struct { uint16_t MinClock; // This is either DCFCLK or SOCCLK (in MHz) uint16_t MaxClock; // This is either DCFCLK or SOCCLK (in MHz)@@ -124,5 +125,6 @@ void dcn315_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr);void dcn315_smu_request_voltage_via_phyclk (struct clk_mgr_internal *clk_mgr,int requested_phyclk_khz);void dcn315_smu_enable_pme_wa (struct clk_mgr_internal *clk_mgr);int dcn315_smu_get_dpref_clk (struct clk_mgr_internal *clk_mgr); -int dcn315_smu_get_smu_fclk (struct clk_mgr_internal *clk_mgr); +int dcn315_smu_get_dtbclk (struct clk_mgr_internal *clk_mgr); +void dcn315_smu_set_dtbclk (struct clk_mgr_internal *clk_mgr,bool enable);#endif /* DAL_DC_315_SMU_H_ */
Kind regards,
Paul
PATCH 6/13 drm/amd/display: Update VTEM Infopacket definition
In reply to Alex HungFrom: "Leo (Hanghong) Ma" <[email protected]>
[Why & How]
The latest HDMI SPEC has updated the VTEM packet structure, so change the VTEM Infopacket defined in the driver side to align with the SPEC.
--- .../gpu/drm/amd/display/modules/info_packet/info_packet.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c +3 -2
@@ -100,7 +100,8 @@ enum vsc_packet_revision {//PB7 = MD0 #define MASK_VTEM_MD0__VRR_EN 0x01#define MASK_VTEM_MD0__M_CONST 0x02 -#define MASK_VTEM_MD0__RESERVED2 0x0C +#define MASK_VTEM_MD0__QMS_EN 0x04 +#define MASK_VTEM_MD0__RESERVED2 0x08#define MASK_VTEM_MD0__FVA_FACTOR_M1 0xF0//MD1
@@ -109,7 +110,7 @@ enum vsc_packet_revision {//MD2 #define MASK_VTEM_MD2__BASE_REFRESH_RATE_98 0x03#define MASK_VTEM_MD2__RB 0x04 -#define MASK_VTEM_MD2__RESERVED3 0xF8 +#define MASK_VTEM_MD2__NEXT_TFR 0xF8//MD3 #define MASK_VTEM_MD3__BASE_REFRESH_RATE_07 0xFF
Signature · 1 line
2.35.1
Re: [PATCH 06/13] drm/amd/display: Update VTEM Infopacket definition
Dear Alex, dear Leo,
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 4 lines
From: "Leo (Hanghong) Ma" <[email protected]>
[Why & How]
The latest HDMI SPEC has updated the VTEM packet structure,
Please add the exact version, as “latest” is relative.
Quoted 2 lines
so change the VTEM Infopacket defined in the driver side to align with the SPEC.
Kind regards,
Paul
Quoted 30 lines
Reviewed-by Chris Park <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Leo (Hanghong) Ma <[email protected]>--- .../gpu/drm/amd/display/modules/info_packet/info_packet.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c +3
-2@@ -100,7 +100,8 @@ enum vsc_packet_revision {//PB7 = MD0 #define MASK_VTEM_MD0__VRR_EN 0x01#define MASK_VTEM_MD0__M_CONST 0x02 -#define MASK_VTEM_MD0__RESERVED2 0x0C +#define MASK_VTEM_MD0__QMS_EN 0x04 +#define MASK_VTEM_MD0__RESERVED2 0x08#define MASK_VTEM_MD0__FVA_FACTOR_M1 0xF0//MD1 @@ -109,7 +110,7 @@ enum vsc_packet_revision {//MD2 #define MASK_VTEM_MD2__BASE_REFRESH_RATE_98 0x03#define MASK_VTEM_MD2__RB 0x04 -#define MASK_VTEM_MD2__RESERVED3 0xF8 +#define MASK_VTEM_MD2__NEXT_TFR 0xF8//MD3 #define MASK_VTEM_MD3__BASE_REFRESH_RATE_07 0xFF
PATCH 7/13 drm/amd/display: [FW Promotion] Release 0.0.109.0
In reply to Alex HungFrom: Anthony Koo <[email protected]>
--- drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +2 -2
@@ -46,10 +46,10 @@/* Firmware versioning. */ #ifdef DMUB_EXPOSE_VERSION -#define DMUB_FW_VERSION_GIT_HASH 0x929554ba +#define DMUB_FW_VERSION_GIT_HASH 0x51b95a35#define DMUB_FW_VERSION_MAJOR 0#define DMUB_FW_VERSION_MINOR 0 -#define DMUB_FW_VERSION_REVISION 108 +#define DMUB_FW_VERSION_REVISION 109#define DMUB_FW_VERSION_TEST 0#define DMUB_FW_VERSION_VBIOS 0#define DMUB_FW_VERSION_HOTFIX 0
Signature · 1 line
2.35.1
Re: [PATCH 07/13] drm/amd/display: [FW Promotion] Release 0.0.109.0
Dear Alex, dear Anthony,
Thank you for the patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 1 line
From: Anthony Koo <[email protected]>
Without a commit message a person not involved in AMD graphics development has no idea, what this version bump is useful for, and what “FW Promotion” is. Could you please add one, and also do it in the future?
Quoted 12 lines
Reviewed-by Aric Cyr <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Anthony Koo <[email protected]>--- drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +0
-0
drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h line 49
Quoted 5 lines
drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +1
-1@@ -46,10 +46,10 @@/* Firmware versioning. */ #ifdef DMUB_EXPOSE_VERSION -#define DMUB_FW_VERSION_GIT_HASH 0x929554ba +#define DMUB_FW_VERSION_GIT_HASH 0x51b95a35
Unrelated to this patch, but could you please add a comment to what git archive this hash refers to?
drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h line 55
Quoted 7 lines
drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +1
-1#define DMUB_FW_VERSION_MAJOR 0#define DMUB_FW_VERSION_MINOR 0 -#define DMUB_FW_VERSION_REVISION 108 +#define DMUB_FW_VERSION_REVISION 109#define DMUB_FW_VERSION_TEST 0#define DMUB_FW_VERSION_VBIOS 0#define DMUB_FW_VERSION_HOTFIX 0
Kind regards,
Paul
PATCH 8/13 drm/amd/display: 3.2.178
In reply to Alex HungFrom: Aric Cyr <[email protected]>
This version brings along following fixes:
- HDCP SEND AKI INIT error
- fix audio format not updated after edid updated
- Reduce stack size
- FEC check in timing validation
- Add fSMC_MSG_SetDtbClk support
- Update VTEM Infopacket definition
- [FW Promotion] Release 0.0.109.0
--- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/dc.h +1 -1
@@ -47,7 +47,7 @@ struct aux_payload;struct set_config_cmd_payload ;struct dmub_notification ; -#define DC_VER "3.2.177" +#define DC_VER "3.2.178"#define MAX_SURFACES 3#define MAX_PLANES 6
Signature · 1 line
2.35.1
PATCH 9/13 drm/amd/display: Add support for zstate during extended vblank
In reply to Alex HungFrom: Gabe Teeger <[email protected]>
[why]
When we enter FREESYNC_STATE_VIDEO, we want to use the extra vblank portion to enter zstate if possible.
[how]
When we enter freesync, a full update is triggered and the new vtotal with extra lines is passed to dml in a stream update. The time gained from extra vblank lines is calculated in microseconds. We allow zstate entry if the time gained is greater than 5 ms, which is the current policy. Furthermore, an optimized value for min_dst_y_next_start is calculated and written to its register. When exiting freesync, another full update is triggered and default values are restored.
--- drivers/gpu/drm/amd/display/dc/core/dc.c | 19 +++++++++++++++++++ drivers/gpu/drm/amd/display/dc/dc.h | 6 +++++- drivers/gpu/drm/amd/display/dc/dc_stream.h | 2 ++ .../drm/amd/display/dc/dcn20/dcn20_hwseq.c | 12 ++++++++++++ .../gpu/drm/amd/display/dc/dcn31/dcn31_hubp.c | 8 ++++++++ .../drm/amd/display/dc/dcn31/dcn31_resource.c | 1 + .../drm/amd/display/dc/dml/dcn20/dcn20_fpu.c | 18 +++++++++++++++--- .../dc/dml/dcn31/display_rq_dlg_calc_31.c | 13 +++++++++++++ .../amd/display/dc/dml/display_mode_structs.h | 2 ++ drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h | 3 +++ 10 files changed, 80 insertions(+), 4 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc.c +19 -0
@@ -2393,6 +2393,8 @@ static enum surface_update_type check_update_surfaces_for_stream(if (stream_update->mst_bw_update) su_flags->bits.mst_bw =1 ; +if (stream_update->crtc_timing_adjust &&dc_extended_blank_supported (dc)) + su_flags->bits.crtc_timing_adjust =1 ;if (su_flags->raw !=0 ) overall_type =UPDATE_TYPE_FULL ;
@@ -2654,6 +2656,9 @@ static void copy_stream_update_to_stream(struct dc *dc,if (update->vrr_infopacket) stream->vrr_infopacket = *update->vrr_infopacket; +if (update->crtc_timing_adjust) + stream->adjust = *update->crtc_timing_adjust; +if (update->dpms_off) stream->dpms_off = *update->dpms_off;
@@ -4055,3 +4060,17 @@ void dc_notify_vsync_int_state(struct dc *dc, struct dc_stream_state *stream, boif (pipe->stream_res.abm && pipe->stream_res.abm->funcs->set_abm_pause) pipe->stream_res.abm->funcs->set_abm_pause (pipe->stream_res.abm, !enable, i, pipe->stream_res.tg->inst); } +/* +* dc_extended_blank_supported: Decide whether extended blank is supported +* +* Extended blank is a freesync optimization feature to be enabled in the future. +* During the extra vblank period gained from freesync, we have the ability to enter z9/z10. +* +* @param [in] dc: Current DC state +* @return: Indicate whether extended blank is supported (true or false) +*/ +bool dc_extended_blank_supported (struct dc *dc) +{ +return dc->debug.extended_blank_optimization && !dc->debug.disable_z10 + && dc->caps.zstate_support && dc->caps.is_apu; +}
drivers/gpu/drm/amd/display/dc/dc.h +5 -1
@@ -188,6 +188,7 @@ struct dc_caps {bool psp_setup_panel_mode;bool extended_aux_timeout_support;bool dmcub_support; +bool zstate_support;uint32_t num_of_internal_disp;enum dp_protocol_version max_dp_protocol_version;unsigned int mall_size_per_mem_channel;
@@ -703,13 +704,14 @@ struct dc_debug_options {bool enable_driver_sequence_debug;enum det_size crb_alloc_policy;int crb_alloc_policy_min_disp_count; -#if defined(CONFIG_DRM_AMD_DC_DCN )bool disable_z10; +#if defined(CONFIG_DRM_AMD_DC_DCN )bool enable_z9_disable_interface;bool enable_sw_cntl_psr;union dpia_debug_options dpia_debug;#endif bool apply_vendor_specific_lttpr_wa; +bool extended_blank_optimization;bool ignore_dpref_ss;uint8_t psr_power_use_phy_fsm; };
@@ -1369,6 +1371,8 @@ struct dc_sink_init_data {bool converter_disable_audio; }; +bool dc_extended_blank_supported (struct dc *dc); +struct dc_sink *dc_sink_create (const struct dc_sink_init_data *init_params);/* Newer interfaces */
drivers/gpu/drm/amd/display/dc/dc_stream.h +2 -0
@@ -131,6 +131,7 @@ union stream_update_flags { uint32_t wb_update:1; uint32_t dsc_changed : 1; uint32_t mst_bw : 1; + uint32_t crtc_timing_adjust : 1; } bits;uint32_t raw;
@@ -289,6 +290,7 @@ struct dc_stream_update {struct dc_3dlut *lut3d_func;struct test_pattern *pending_test_pattern; +struct dc_crtc_timing_adjust *crtc_timing_adjust; }; bool dc_is_stream_unchanged(
drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +12 -0
@@ -1857,6 +1857,7 @@ void dcn20_optimize_bandwidth( struct dc_state *context) {struct hubbub *hubbub = dc->res_pool->hubbub; +int i;/* program dchubbub watermarks */ hubbub->funcs->program_watermarks(hubbub,
@@ -1873,6 +1874,17 @@ void dcn20_optimize_bandwidth( dc->clk_mgr, context, true); +if (dc_extended_blank_supported (dc) && context->bw_ctx.bw.dcn.clk.zstate_support ==DCN_ZSTATE_SUPPORT_ALLOW ) { +for (i =0 ; i < dc->res_pool->pipe_count; ++i) { +struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; + +if (pipe_ctx->stream && pipe_ctx->plane_res.hubp->funcs->program_extended_blank + && pipe_ctx->stream->adjust.v_total_min == pipe_ctx->stream->adjust.v_total_max + && pipe_ctx->stream->adjust.v_total_max > pipe_ctx->stream->timing.v_total) + pipe_ctx->plane_res.hubp->funcs->program_extended_blank (pipe_ctx->plane_res.hubp, + pipe_ctx->dlg_regs.optimized_min_dst_y_next_start); + } + }/* increase compbuf size */ if (hubbub->funcs->program_compbuf_size) hubbub->funcs->program_compbuf_size (hubbub, context->bw_ctx.bw.dcn.compbuf_size_kb, true);
drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubp.c +8 -0
@@ -54,6 +54,13 @@ void hubp31_soft_reset(struct hubp *hubp, bool reset)REG_UPDATE (DCHUBP_CNTL ,HUBP_SOFT_RESET , reset); } +void hubp31_program_extended_blank (struct hubp *hubp,unsigned int min_dst_y_next_start_optimized) +{ +struct dcn20_hubp *hubp2 =TO_DCN20_HUBP (hubp); + +REG_SET (BLANK_OFFSET_1 ,0 ,MIN_DST_Y_NEXT_START , min_dst_y_next_start_optimized); +} +static struct hubp_funcs dcn31_hubp_funcs = { .hubp_enable_tripleBuffer = hubp2_enable_triplebuffer, .hubp_is_triplebuffer_enabled = hubp2_is_triplebuffer_enabled,
@@ -80,6 +87,7 @@ static struct hubp_funcs dcn31_hubp_funcs = { .set_unbounded_requesting = hubp31_set_unbounded_requesting, .hubp_soft_reset = hubp31_soft_reset, .hubp_in_blank = hubp1_in_blank, + .program_extended_blank = hubp31_program_extended_blank, }; bool hubp31_construct(
drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +1 -0
@@ -2232,6 +2232,7 @@ static bool dcn31_resource_construct( dc->caps.extended_aux_timeout_support = true; dc->caps.dmcub_support = true; dc->caps.is_apu = true; + dc->caps.zstate_support = true;/* Color pipeline capabilities */ dc->caps.color.dpp.dcn_arch =1 ;
drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c +15 -3
@@ -722,8 +722,10 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, struc {int plane_count;int i; +unsigned int optimized_min_dst_y_next_start_us; plane_count =0 ; + optimized_min_dst_y_next_start_us =0 ;for (i =0 ; i < dc->res_pool->pipe_count; i++) {if (context->res_ctx.pipe_ctx[i].plane_state) plane_count++;
@@ -744,11 +746,22 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, strucstruct dc_link *link = context->streams[0 ]->sink->link;struct dc_stream_status *stream_status = &context->stream_status[0 ]; +if (dc_extended_blank_supported (dc)) { +for (i =0 ; i < dc->res_pool->pipe_count; i++) { +if (context->res_ctx.pipe_ctx[i].stream == context->streams[0 ] + && context->res_ctx.pipe_ctx[i].stream->adjust.v_total_min == context->res_ctx.pipe_ctx[i].stream->adjust.v_total_max + && context->res_ctx.pipe_ctx[i].stream->adjust.v_total_min > context->res_ctx.pipe_ctx[i].stream->timing.v_total) { + optimized_min_dst_y_next_start_us = + context->res_ctx.pipe_ctx[i].dlg_regs.optimized_min_dst_y_next_start_us; +break ; + } + } + }/* zstate only supported on PWRSEQ0 and when there's <2 planes*/ if (link->link_index !=0 || stream_status->plane_count >1 )return DCN_ZSTATE_SUPPORT_DISALLOW ; -if (context->bw_ctx.dml.vba.StutterPeriod >5000.0 ) +if (context->bw_ctx.dml.vba.StutterPeriod >5000.0 || optimized_min_dst_y_next_start_us >5000 )return DCN_ZSTATE_SUPPORT_ALLOW ;else if (link->psr_settings.psr_version ==DC_PSR_VERSION_1 && !dc->debug.disable_psr)return DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY ;
@@ -786,8 +799,6 @@ void dcn20_calculate_dlg_params( != dm_dram_clock_change_unsupported; context->bw_ctx.bw.dcn.clk.dppclk_khz =0 ; - context->bw_ctx.bw.dcn.clk.zstate_support =decide_zstate_support (dc, context); - context->bw_ctx.bw.dcn.clk.dtbclk_en =is_dtbclk_required (dc, context);if (context->bw_ctx.bw.dcn.clk.dispclk_khz < dc->debug.min_disp_clk_khz)
@@ -843,6 +854,7 @@ void dcn20_calculate_dlg_params( &pipes[pipe_idx].pipe); pipe_idx++; } + context->bw_ctx.bw.dcn.clk.zstate_support =decide_zstate_support (dc, context); } static void swizzle_to_dml_params(
drivers/gpu/drm/amd/display/dc/dml/dcn31/display_rq_dlg_calc_31.c +13 -0
@@ -1055,6 +1055,7 @@ static void dml_rq_dlg_get_dlg_params(float vba__refcyc_per_req_delivery_pre_l =get_refcyc_per_req_delivery_pre_l_in_us (mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz;// From VBA float vba__refcyc_per_req_delivery_l =get_refcyc_per_req_delivery_l_in_us (mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz;// From VBA +int blank_lines;memset (disp_dlg_regs,0 ,sizeof (*disp_dlg_regs));memset (disp_ttu_regs,0 ,sizeof (*disp_ttu_regs));
@@ -1080,6 +1081,18 @@ static void dml_rq_dlg_get_dlg_params( dlg_vblank_start = interlaced ? (vblank_start /2 ) : vblank_start; disp_dlg_regs->min_dst_y_next_start = (unsigned int ) (((double ) dlg_vblank_start) *dml_pow (2 ,2 )); + blank_lines = (dst->vblank_end + dst->vtotal_min - dst->vblank_start - dst->vstartup_start -1 ); +if (blank_lines <0 ) + blank_lines =0 ; +if (blank_lines !=0 ) { + disp_dlg_regs->optimized_min_dst_y_next_start_us = + ((unsigned int ) blank_lines * dst->hactive) / (unsigned int ) dst->pixel_rate_mhz; + disp_dlg_regs->optimized_min_dst_y_next_start = + (unsigned int )(((double ) (dlg_vblank_start + blank_lines)) *dml_pow (2 ,2 )); + }else { +// use unoptimized value + disp_dlg_regs->optimized_min_dst_y_next_start = disp_dlg_regs->min_dst_y_next_start; + }ASSERT (disp_dlg_regs->min_dst_y_next_start < (unsigned int )dml_pow (2 ,18 ));dml_print ("DML_DLG: %s: min_ttu_vblank (us) = %3.2f\n" , __func__, min_ttu_vblank);
drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h +2 -0
@@ -446,6 +446,8 @@ struct _vcs_dpi_display_dlg_regs_st {unsigned int refcyc_h_blank_end;unsigned int dlg_vblank_end;unsigned int min_dst_y_next_start; +unsigned int optimized_min_dst_y_next_start; +unsigned int optimized_min_dst_y_next_start_us;unsigned int refcyc_per_htotal;unsigned int refcyc_x_after_scaler;unsigned int dst_y_after_scaler;
drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h +3 -0
@@ -195,6 +195,9 @@ struct hubp_funcs {void (*hubp_set_flip_int)(struct hubp *hubp); +void (*program_extended_blank)(struct hubp *hubp, +unsigned int min_dst_y_next_start_optimized); +void (*hubp_wait_pipe_read_start)(struct hubp *hubp); };
Signature · 1 line
2.35.1
PATCH 10/13 drm/amd/display: remove destructive verify link for TMDS
In reply to Alex HungFrom: Charlene Liu <[email protected]>
[why and how]
TMDS not need destructive verify link
--- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc_link.c +1 -2
@@ -983,8 +983,7 @@ static bool should_verify_link_capability_destructively(struct dc_link *link, destrictive = false; } } - }else if (dc_is_hdmi_signal (link->local_sink->sink_signal)) - destrictive = true; + }return destrictive; }
Signature · 1 line
2.35.1
Re: [PATCH 10/13] drm/amd/display: remove destructive verify link for TMDS
Dear Alex, dear Charlene,
Thank you for the patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 4 lines
From: Charlene Liu <[email protected]>
[why and how]
TMDS not need destructive verify link
According to whom? Please document the source for that claim.
Is that just to be spec compliant, or does it cause an actual problem? If so, how can it be reproduced?
Quoted 13 lines
Reviewed-by Aric Cyr <[email protected]>Acked-by Alan Liu <[email protected]>Acked-by Alex Hung <[email protected]>Signed-off-by Charlene Liu <[email protected]>--- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)drivers/gpu/drm/amd/display/dc/core/dc_link.c +0
-0
drivers/gpu/drm/amd/display/dc/core/dc_link.c static bool should_verify_link_capability_destructively(struct dc_link *link, · line 989
Quoted 9 lines
drivers/gpu/drm/amd/display/dc/core/dc_link.c +1
-2@@ -983,8 +983,7 @@ static bool should_verify_link_capability_destructively(struct dc_link *link, destrictive = false; } } - }else if (dc_is_hdmi_signal (link->local_sink->sink_signal)) - destrictive = true; + }return destrictive; }
It’d be great, if you also sent a patch to fix the typo in the variable name.
Kind regards,
Paul
PATCH 11/13 drm/amd/dicplay: move FPU related code from dcn31 to dml/dcn31 folder
In reply to Alex HungFrom: Melissa Wen <[email protected]>
Creates FPU files in dml/dcn31 folder to centralize FPU operations from 3.1x drivers and moves all FPU-associated code from dcn31 driver to there. It includes the struct _vcs_dpi_ip_params_st and _vcs_dpi_soc_bounding_box_st and functions:
- dcn31_calculate_wm_and_dlg_fp()
- dcn31_update_bw_bounding_box()
adding dc_assert_fp_enabled to them and drop DC_FP_START/END inside functions that was moved to dml folder, as required.
--- drivers/gpu/drm/amd/display/dc/dcn31/Makefile | 26 -- .../drm/amd/display/dc/dcn31/dcn31_resource.c | 355 +-------------- .../drm/amd/display/dc/dcn31/dcn31_resource.h | 4 +- drivers/gpu/drm/amd/display/dc/dml/Makefile | 2 + .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 406 ++++++++++++++++++ .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 39 ++ 6 files changed, 451 insertions(+), 381 deletions(-) create mode 100644 drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c create mode 100644 drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h
drivers/gpu/drm/amd/display/dc/dcn31/Makefile +0 -26
@@ -15,32 +15,6 @@ DCN31 = dcn31_resource.o dcn31_hubbub.o dcn31_hwseq.o dcn31_init.o dcn31_hubp.o dcn31_apg.o dcn31_hpo_dp_stream_encoder.o dcn31_hpo_dp_link_encoder.o \ dcn31_afmt.o dcn31_vpg.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -msse2 -endif -endif - AMD_DAL_DCN31 = $(addprefix $(AMDDALPATH)/dc/dcn31/,$(DCN31)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN31)
drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +1 -354
379 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h +3 -1
@@ -31,6 +31,9 @@#define TO_DCN31_RES_POOL (pool)\ container_of(pool, struct dcn31_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_1_ip; +extern struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc; +struct dcn31_resource_pool {struct resource_pool base; };
@@ -47,7 +50,6 @@ int dcn31_populate_dml_pipes_from_context( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, bool fast_validate); -void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params);void dcn31_update_soc_for_wm_a (struct dc *dc,struct dc_state *context); struct resource_pool *dcn31_create_resource_pool(
drivers/gpu/drm/amd/display/dc/dml/Makefile +2 -0
@@ -71,6 +71,7 @@ CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_mode_vba_30.o := $(dml_ccflags) $(fram CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_rq_dlg_calc_30.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_mode_vba_31.o := $(dml_ccflags) $(frame_warn_flag) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_rq_dlg_calc_31.o := $(dml_ccflags) +CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_ccflags)
@@ -114,6 +115,7 @@ DML += dcn20/display_rq_dlg_calc_20v2.o dcn20/display_mode_vba_20v2.o DML += dcn21/display_rq_dlg_calc_21.o dcn21/display_mode_vba_21.o DML += dcn30/display_mode_vba_30.o dcn30/display_rq_dlg_calc_30.o DML += dcn31/display_mode_vba_31.o dcn31/display_rq_dlg_calc_31.o +DML += dcn31/dcn31_fpu.o DML += dcn301/dcn301_fpu.o DML += dcn302/dcn302_fpu.o DML += dcn303/dcn303_fpu.o
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c new file +406 -0
406 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h new file +39 -0
@@ -0,0 +1,39 @@ +/* +* Copyright 2019-2021 Advanced Micro Devices, Inc. +* +* Permission is hereby granted, free of charge, to any person obtaining a +* copy of this software and associated documentation files (the "Software"), +* to deal in the Software without restriction, including without limitation +* the rights to use, copy, modify, merge, publish, distribute, sublicense, +* and/or sell copies of the Software, and to permit persons to whom the +* Software is furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +* OTHER DEALINGS IN THE SOFTWARE. +* +* Authors: AMD +* +*/ + +#ifndef __DCN31_FPU_H__ +#define __DCN31_FPU_H__ + +#define DCN3_1_DEFAULT_DET_SIZE 384 + +void dcn31_calculate_wm_and_dlg_fp ( +struct dc *dc,struct dc_state *context, +display_e2e_pipe_params_st *pipes, +int pipe_cnt, +int vlevel); + +void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params); + +#endif /* __DCN31_FPU_H__*/
Signature · 1 line
2.35.1
Re: [PATCH 11/13] drm/amd/dicplay: move FPU related code from dcn31 to dml/dcn31 folder
Dear Alex, dear Melissa,
Thank you for the patch.
Am 18.03.22 um 22:47 schrieb Alex Hung:
Quoted 1 line
From: Melissa Wen <[email protected]>
Typo in the commit message summary: di*s*play
Quoted 10 lines
Creates FPU files in dml/dcn31 folder to centralize FPU operations from 3.1x drivers and moves all FPU-associated code from dcn31 driver to there. It includes the struct
_vcs_dpi_ip_params_stand_vcs_dpi_soc_bounding_box_stand functions:-
dcn31_calculate_wm_and_dlg_fp()
-dcn31_update_bw_bounding_box()adding
dc_assert_fp_enabledto them and drop DC_FP_START/END inside functions that was moved to dml folder, as required.
In what commit was it moved?
Is there an easy way to verify, this move had no sideeffects? (If it passes the build bots, everything is fine?)
Kind regards,
Paul
Quoted 936 lines
Signed-off-by Melissa Wen <[email protected]>Reviewed-by Alex Hung <[email protected]>--- drivers/gpu/drm/amd/display/dc/dcn31/Makefile | 26 -- .../drm/amd/display/dc/dcn31/dcn31_resource.c | 355 +-------------- .../drm/amd/display/dc/dcn31/dcn31_resource.h | 4 +- drivers/gpu/drm/amd/display/dc/dml/Makefile | 2 + .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 406 ++++++++++++++++++ .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 39 ++ 6 files changed, 451 insertions(+), 381 deletions(-)create mode 100644
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.ccreate mode 100644drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.hdrivers/gpu/drm/amd/display/dc/dcn31/Makefile +0
-26@@ -15,32 +15,6 @@ DCN31 = dcn31_resource.o dcn31_hubbub.o dcn31_hwseq.o dcn31_init.o dcn31_hubp.o dcn31_apg.o dcn31_hpo_dp_stream_encoder.o dcn31_hpo_dp_link_encoder.o \ dcn31_afmt.o dcn31_vpg.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -msse2 -endif -endif - AMD_DAL_DCN31 = $(addprefix $(AMDDALPATH)/dc/dcn31/,$(DCN31)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN31)drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +1
-354@@ -65,6 +65,7 @@#include "virtual/virtual_stream_encoder.h" #include "dce110/dce110_resource.h" #include "dml/display_mode_vba.h" +#include "dml/dcn31/dcn31_fpu.h" #include "dcn31/dcn31_dccg.h" #include "dcn10/dcn10_resource.h" #include "dcn31_panel_cntl.h" @@ -102,152 +103,6 @@#define DC_LOGGER_INIT (logger) -#define DCN3_1_DEFAULT_DET_SIZE 384 - -struct _vcs_dpi_ip_params_st dcn3_1_ip = { - .gpuvm_enable =1 , - .gpuvm_max_page_table_levels =1 , - .hostvm_enable =1 , - .hostvm_max_page_table_levels =2 , - .rob_buffer_size_kbytes =64 , - .det_buffer_size_kbytes =DCN3_1_DEFAULT_DET_SIZE , - .config_return_buffer_size_in_kbytes =1792 , - .compressed_buffer_segment_size_in_kbytes =64 , - .meta_fifo_size_in_kentries =32 , - .zero_size_buffer_entries =512 , - .compbuf_reserved_space_64b =256 , - .compbuf_reserved_space_zs =64 , - .dpp_output_buffer_pixels =2560 , - .opp_output_buffer_lines =1 , - .pixel_chunk_size_kbytes =8 , - .meta_chunk_size_kbytes =2 , - .min_meta_chunk_size_bytes =256 , - .writeback_chunk_size_kbytes =8 , - .ptoi_supported = false, - .num_dsc =3 , - .maximum_dsc_bits_per_component =10 , - .dsc422_native_support = false, - .is_line_buffer_bpp_fixed = true, - .line_buffer_fixed_bpp =48 , - .line_buffer_size_bits =789504 , - .max_line_buffer_lines =12 , - .writeback_interface_buffer_size_kbytes =90 , - .max_num_dpp =4 , - .max_num_otg =4 , - .max_num_hdmi_frl_outputs =1 , - .max_num_wb =1 , - .max_dchub_pscl_bw_pix_per_clk =4 , - .max_pscl_lb_bw_pix_per_clk =2 , - .max_lb_vscl_bw_pix_per_clk =4 , - .max_vscl_hscl_bw_pix_per_clk =4 , - .max_hscl_ratio =6 , - .max_vscl_ratio =6 , - .max_hscl_taps =8 , - .max_vscl_taps =8 , - .dpte_buffer_size_in_pte_reqs_luma =64 , - .dpte_buffer_size_in_pte_reqs_chroma =34 , - .dispclk_ramp_margin_percent =1 , - .max_inter_dcn_tile_repeaters =8 , - .cursor_buffer_size =16 , - .cursor_chunk_size =2 , - .writeback_line_buffer_buffer_size =0 , - .writeback_min_hscl_ratio =1 , - .writeback_min_vscl_ratio =1 , - .writeback_max_hscl_ratio =1 , - .writeback_max_vscl_ratio =1 , - .writeback_max_hscl_taps =1 , - .writeback_max_vscl_taps =1 , - .dppclk_delay_subtotal =46 , - .dppclk_delay_scl =50 , - .dppclk_delay_scl_lb_only =16 , - .dppclk_delay_cnvc_formatter =27 , - .dppclk_delay_cnvc_cursor =6 , - .dispclk_delay_subtotal =119 , - .dynamic_metadata_vm_enabled = false, - .odm_combine_4to1_supported = false, - .dcc_supported = true, -}; - -struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc = { -/*TODO: correct dispclk/dppclk voltage level determination*/ - .clock_limits = { - { - .state =0 , - .dispclk_mhz =1200.0 , - .dppclk_mhz =1200.0 , - .phyclk_mhz =600.0 , - .phyclk_d18_mhz =667.0 , - .dscclk_mhz =186.0 , - .dtbclk_mhz =625.0 , - }, - { - .state =1 , - .dispclk_mhz =1200.0 , - .dppclk_mhz =1200.0 , - .phyclk_mhz =810.0 , - .phyclk_d18_mhz =667.0 , - .dscclk_mhz =209.0 , - .dtbclk_mhz =625.0 , - }, - { - .state =2 , - .dispclk_mhz =1200.0 , - .dppclk_mhz =1200.0 , - .phyclk_mhz =810.0 , - .phyclk_d18_mhz =667.0 , - .dscclk_mhz =209.0 , - .dtbclk_mhz =625.0 , - }, - { - .state =3 , - .dispclk_mhz =1200.0 , - .dppclk_mhz =1200.0 , - .phyclk_mhz =810.0 , - .phyclk_d18_mhz =667.0 , - .dscclk_mhz =371.0 , - .dtbclk_mhz =625.0 , - }, - { - .state =4 , - .dispclk_mhz =1200.0 , - .dppclk_mhz =1200.0 , - .phyclk_mhz =810.0 , - .phyclk_d18_mhz =667.0 , - .dscclk_mhz =417.0 , - .dtbclk_mhz =625.0 , - }, - }, - .num_states =5 , - .sr_exit_time_us =9.0 , - .sr_enter_plus_exit_time_us =11.0 , - .sr_exit_z8_time_us =442.0 , - .sr_enter_plus_exit_z8_time_us =560.0 , - .writeback_latency_us =12.0 , - .dram_channel_width_bytes =4 , - .round_trip_ping_latency_dcfclk_cycles =106 , - .urgent_latency_pixel_data_only_us =4.0 , - .urgent_latency_pixel_mixed_with_vm_data_us =4.0 , - .urgent_latency_vm_data_only_us =4.0 , - .urgent_out_of_order_return_per_channel_pixel_only_bytes =4096 , - .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes =4096 , - .urgent_out_of_order_return_per_channel_vm_only_bytes =4096 , - .pct_ideal_sdp_bw_after_urgent =80.0 , - .pct_ideal_dram_sdp_bw_after_urgent_pixel_only =65.0 , - .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm =60.0 , - .pct_ideal_dram_sdp_bw_after_urgent_vm_only =30.0 , - .max_avg_sdp_bw_use_normal_percent =60.0 , - .max_avg_dram_bw_use_normal_percent =60.0 , - .fabric_datapath_to_dcn_data_return_bytes =32 , - .return_bus_width_bytes =64 , - .downspread_percent =0.38 , - .dcn_downspread_percent =0.5 , - .gpuvm_min_page_size_bytes =4096 , - .hostvm_min_page_size_bytes =4096 , - .do_urgent_latency_adjustment = false, - .urgent_latency_adjustment_fabric_clock_component_us =0 , - .urgent_latency_adjustment_fabric_clock_reference_mhz =0 , -}; -enum dcn31_clk_src_array_id {DCN31_CLK_SRC_PLL0 ,DCN31_CLK_SRC_PLL1 ,@@ -1869,143 +1724,6 @@ void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context) } } -static void dcn31_calculate_wm_and_dlg_fp ( -struct dc *dc,struct dc_state *context, -display_e2e_pipe_params_st *pipes, -int pipe_cnt, -int vlevel) -{ -int i, pipe_idx; -double dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; - -if (context->bw_ctx.dml.soc.min_dcfclk > dcfclk) - dcfclk = context->bw_ctx.dml.soc.min_dcfclk; - -/* We don't recalculate clocks for 0 pipe configs, which can block -* S0i3 as high clocks will block low power states -* Override any clocks that can block S0i3 to min here -*/ -if (pipe_cnt ==0 ) { - context->bw_ctx.bw.dcn.clk.dcfclk_khz = dcfclk;// always should be vlevel 0 -return ; - } - - pipes[0 ].clks_cfg.voltage = vlevel; - pipes[0 ].clks_cfg.dcfclk_mhz = dcfclk; - pipes[0 ].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; - -#if 0 // TODO -/* Set B: -* TODO -*/ -if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].valid) { -if (vlevel ==0 ) { - pipes[0 ].clks_cfg.voltage =1 ; - pipes[0 ].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[0 ].dcfclk_mhz; - } - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.b.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.b.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - - pipes[0 ].clks_cfg.voltage = vlevel; - pipes[0 ].clks_cfg.dcfclk_mhz = dcfclk; - -/* Set C: -* TODO -*/ -if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].valid) { - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.c.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.c.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - -/* Set D: -* TODO -*/ -if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].valid) { - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.d.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.d.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; -#endif - -/* Set A: -* All clocks min required -* -* Set A calculated last so that following calculations are based on Set A -*/ - dc->res_pool->funcs->update_soc_for_wm_a (dc, context); - context->bw_ctx.bw.dcn.watermarks.a.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; - context->bw_ctx.bw.dcn.watermarks.a.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; -/* TODO: remove: */ - context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; - context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; - context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; -/* end remove*/ - -for (i =0 , pipe_idx =0 ; i < dc->res_pool->pipe_count; i++) { -if (!context->res_ctx.pipe_ctx[i].stream) -continue ; - - pipes[pipe_idx].clks_cfg.dispclk_mhz =get_dispclk_calculated (&context->bw_ctx.dml, pipes, pipe_cnt); - pipes[pipe_idx].clks_cfg.dppclk_mhz =get_dppclk_calculated (&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx); - -if (dc->config.forced_clocks || dc->debug.max_disp_clk) { - pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0 ].dispclk_mhz; - pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0 ].dppclk_mhz; - } -if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz *1000 ) - pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz /1000.0 ; -if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz *1000 ) - pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz /1000.0 ; - - pipe_idx++; - } - -DC_FP_START (); -dcn20_calculate_dlg_params (dc, context, pipes, pipe_cnt, vlevel); -DC_FP_END (); -} - void dcn31_calculate_wm_and_dlg( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes,@@ -2073,77 +1791,6 @@ static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; -void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params) -{ -struct clk_limit_table *clk_table = &bw_params->clk_table; -struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES ]; -unsigned int i, closest_clk_lvl; -int j; - -// Default clock levels are used for diags, which may lead to overclocking. -if (!IS_DIAG_DC (dc->ctx->dce_environment)) { -int max_dispclk_mhz =0 , max_dppclk_mhz =0 ; - - dcn3_1_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; - dcn3_1_ip.max_num_dpp = dc->res_pool->pipe_count; - dcn3_1_soc.num_chans = bw_params->num_channels; - -ASSERT (clk_table->num_entries); - -/* Prepass to find max clocks independent of voltage level. */ -for (i =0 ; i < clk_table->num_entries; ++i) { -if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) - max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; -if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) - max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; - } - -for (i =0 ; i < clk_table->num_entries; i++) { -/* loop backwards*/ -for (closest_clk_lvl =0 , j = dcn3_1_soc.num_states -1 ; j >=0 ; j--) { -if ((unsigned int ) dcn3_1_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { - closest_clk_lvl = j; -break ; - } - } - - clock_limits[i].state = i; - -/* Clocks dependent on voltage level. */ - clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; - clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; - clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; - clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz *2 * clk_table->entries[i].wck_ratio; - -/* Clocks independent of voltage level. */ - clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : - dcn3_1_soc.clock_limits[closest_clk_lvl].dispclk_mhz; - - clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : - dcn3_1_soc.clock_limits[closest_clk_lvl].dppclk_mhz; - - clock_limits[i].dram_bw_per_chan_gbps = dcn3_1_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; - clock_limits[i].dscclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dscclk_mhz; - clock_limits[i].dtbclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; - clock_limits[i].phyclk_d18_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; - clock_limits[i].phyclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_mhz; - } -for (i =0 ; i < clk_table->num_entries; i++) - dcn3_1_soc.clock_limits[i] = clock_limits[i]; -if (clk_table->num_entries) { - dcn3_1_soc.num_states = clk_table->num_entries; - } - } - - dcn3_1_soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz /1000.0 ; - dc->dml.soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz /1000.0 ; - -if (!IS_FPGA_MAXIMUS_DC (dc->ctx->dce_environment)) -dml_init_instance (&dc->dml, &dcn3_1_soc, &dcn3_1_ip,DML_PROJECT_DCN31 ); -else -dml_init_instance (&dc->dml, &dcn3_1_soc, &dcn3_1_ip,DML_PROJECT_DCN31_FPGA ); -} -static struct resource_funcs dcn31_res_pool_funcs = { .destroy = dcn31_destroy_resource_pool, .link_enc_create = dcn31_link_encoder_create,drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h +3
-1@@ -31,6 +31,9 @@#define TO_DCN31_RES_POOL (pool)\ container_of(pool, struct dcn31_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_1_ip; +extern struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc; +struct dcn31_resource_pool {struct resource_pool base; };@@ -47,7 +50,6 @@ int dcn31_populate_dml_pipes_from_context( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, bool fast_validate); -void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params);void dcn31_update_soc_for_wm_a (struct dc *dc,struct dc_state *context); struct resource_pool *dcn31_create_resource_pool(drivers/gpu/drm/amd/display/dc/dml/Makefile +2
-0@@ -71,6 +71,7 @@ CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_mode_vba_30.o := $(dml_ccflags) $(fram CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_rq_dlg_calc_30.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_mode_vba_31.o := $(dml_ccflags) $(frame_warn_flag) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_rq_dlg_calc_31.o := $(dml_ccflags) +CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_ccflags)@@ -114,6 +115,7 @@ DML += dcn20/display_rq_dlg_calc_20v2.o dcn20/display_mode_vba_20v2.o DML += dcn21/display_rq_dlg_calc_21.o dcn21/display_mode_vba_21.o DML += dcn30/display_mode_vba_30.o dcn30/display_rq_dlg_calc_30.o DML += dcn31/display_mode_vba_31.o dcn31/display_rq_dlg_calc_31.o +DML += dcn31/dcn31_fpu.o DML += dcn301/dcn301_fpu.o DML += dcn302/dcn302_fpu.o DML += dcn303/dcn303_fpu.odrivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c new file +406
-0@@ -0,0 +1,406 @@ +/* +* Copyright 2019-2021 Advanced Micro Devices, Inc. +* +* Permission is hereby granted, free of charge, to any person obtaining a +* copy of this software and associated documentation files (the "Software"), +* to deal in the Software without restriction, including without limitation +* the rights to use, copy, modify, merge, publish, distribute, sublicense, +* and/or sell copies of the Software, and to permit persons to whom the +* Software is furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +* OTHER DEALINGS IN THE SOFTWARE. +* +* Authors: AMD +* +*/ + +#include "resource.h" +#include "clk_mgr.h" + +#include "dml/dcn20/dcn20_fpu.h" +#include "dcn31_fpu.h" + +/** +* DOC: DCN31x FPU manipulation Overview +* +* The DCN architecture relies on FPU operations, which require special +* compilation flags and the use of kernel_fpu_begin/end functions; ideally, we +* want to avoid spreading FPU access across multiple files. With this idea in +* mind, this file aims to centralize all DCN3.1.x functions that require FPU +* access in a single place. Code in this file follows the following code +* pattern: +* +* 1. Functions that use FPU operations should be isolated in static functions. +* 2. The FPU functions should have the noinline attribute to ensure anything +* that deals with FP register is contained within this call. +* 3. All function that needs to be accessed outside this file requires a +* public interface that not uses any FPU reference. +* 4. Developers **must not** use DC_FP_START/END in this file, but they need +* to ensure that the caller invokes it before access any function available +* in this file. For this reason, public functions in this file must invoke +* dc_assert_fp_enabled(); +*/ + +struct _vcs_dpi_ip_params_st dcn3_1_ip = { + .gpuvm_enable =1 , + .gpuvm_max_page_table_levels =1 , + .hostvm_enable =1 , + .hostvm_max_page_table_levels =2 , + .rob_buffer_size_kbytes =64 , + .det_buffer_size_kbytes =DCN3_1_DEFAULT_DET_SIZE , + .config_return_buffer_size_in_kbytes =1792 , + .compressed_buffer_segment_size_in_kbytes =64 , + .meta_fifo_size_in_kentries =32 , + .zero_size_buffer_entries =512 , + .compbuf_reserved_space_64b =256 , + .compbuf_reserved_space_zs =64 , + .dpp_output_buffer_pixels =2560 , + .opp_output_buffer_lines =1 , + .pixel_chunk_size_kbytes =8 , + .meta_chunk_size_kbytes =2 , + .min_meta_chunk_size_bytes =256 , + .writeback_chunk_size_kbytes =8 , + .ptoi_supported = false, + .num_dsc =3 , + .maximum_dsc_bits_per_component =10 , + .dsc422_native_support = false, + .is_line_buffer_bpp_fixed = true, + .line_buffer_fixed_bpp =48 , + .line_buffer_size_bits =789504 , + .max_line_buffer_lines =12 , + .writeback_interface_buffer_size_kbytes =90 , + .max_num_dpp =4 , + .max_num_otg =4 , + .max_num_hdmi_frl_outputs =1 , + .max_num_wb =1 , + .max_dchub_pscl_bw_pix_per_clk =4 , + .max_pscl_lb_bw_pix_per_clk =2 , + .max_lb_vscl_bw_pix_per_clk =4 , + .max_vscl_hscl_bw_pix_per_clk =4 , + .max_hscl_ratio =6 , + .max_vscl_ratio =6 , + .max_hscl_taps =8 , + .max_vscl_taps =8 , + .dpte_buffer_size_in_pte_reqs_luma =64 , + .dpte_buffer_size_in_pte_reqs_chroma =34 , + .dispclk_ramp_margin_percent =1 , + .max_inter_dcn_tile_repeaters =8 , + .cursor_buffer_size =16 , + .cursor_chunk_size =2 , + .writeback_line_buffer_buffer_size =0 , + .writeback_min_hscl_ratio =1 , + .writeback_min_vscl_ratio =1 , + .writeback_max_hscl_ratio =1 , + .writeback_max_vscl_ratio =1 , + .writeback_max_hscl_taps =1 , + .writeback_max_vscl_taps =1 , + .dppclk_delay_subtotal =46 , + .dppclk_delay_scl =50 , + .dppclk_delay_scl_lb_only =16 , + .dppclk_delay_cnvc_formatter =27 , + .dppclk_delay_cnvc_cursor =6 , + .dispclk_delay_subtotal =119 , + .dynamic_metadata_vm_enabled = false, + .odm_combine_4to1_supported = false, + .dcc_supported = true, +}; + +struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc = { +/*TODO: correct dispclk/dppclk voltage level determination*/ + .clock_limits = { + { + .state =0 , + .dispclk_mhz =1200.0 , + .dppclk_mhz =1200.0 , + .phyclk_mhz =600.0 , + .phyclk_d18_mhz =667.0 , + .dscclk_mhz =186.0 , + .dtbclk_mhz =625.0 , + }, + { + .state =1 , + .dispclk_mhz =1200.0 , + .dppclk_mhz =1200.0 , + .phyclk_mhz =810.0 , + .phyclk_d18_mhz =667.0 , + .dscclk_mhz =209.0 , + .dtbclk_mhz =625.0 , + }, + { + .state =2 , + .dispclk_mhz =1200.0 , + .dppclk_mhz =1200.0 , + .phyclk_mhz =810.0 , + .phyclk_d18_mhz =667.0 , + .dscclk_mhz =209.0 , + .dtbclk_mhz =625.0 , + }, + { + .state =3 , + .dispclk_mhz =1200.0 , + .dppclk_mhz =1200.0 , + .phyclk_mhz =810.0 , + .phyclk_d18_mhz =667.0 , + .dscclk_mhz =371.0 , + .dtbclk_mhz =625.0 , + }, + { + .state =4 , + .dispclk_mhz =1200.0 , + .dppclk_mhz =1200.0 , + .phyclk_mhz =810.0 , + .phyclk_d18_mhz =667.0 , + .dscclk_mhz =417.0 , + .dtbclk_mhz =625.0 , + }, + }, + .num_states =5 , + .sr_exit_time_us =9.0 , + .sr_enter_plus_exit_time_us =11.0 , + .sr_exit_z8_time_us =442.0 , + .sr_enter_plus_exit_z8_time_us =560.0 , + .writeback_latency_us =12.0 , + .dram_channel_width_bytes =4 , + .round_trip_ping_latency_dcfclk_cycles =106 , + .urgent_latency_pixel_data_only_us =4.0 , + .urgent_latency_pixel_mixed_with_vm_data_us =4.0 , + .urgent_latency_vm_data_only_us =4.0 , + .urgent_out_of_order_return_per_channel_pixel_only_bytes =4096 , + .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes =4096 , + .urgent_out_of_order_return_per_channel_vm_only_bytes =4096 , + .pct_ideal_sdp_bw_after_urgent =80.0 , + .pct_ideal_dram_sdp_bw_after_urgent_pixel_only =65.0 , + .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm =60.0 , + .pct_ideal_dram_sdp_bw_after_urgent_vm_only =30.0 , + .max_avg_sdp_bw_use_normal_percent =60.0 , + .max_avg_dram_bw_use_normal_percent =60.0 , + .fabric_datapath_to_dcn_data_return_bytes =32 , + .return_bus_width_bytes =64 , + .downspread_percent =0.38 , + .dcn_downspread_percent =0.5 , + .gpuvm_min_page_size_bytes =4096 , + .hostvm_min_page_size_bytes =4096 , + .do_urgent_latency_adjustment = false, + .urgent_latency_adjustment_fabric_clock_component_us =0 , + .urgent_latency_adjustment_fabric_clock_reference_mhz =0 , +}; + + +void dcn31_calculate_wm_and_dlg_fp ( +struct dc *dc,struct dc_state *context, +display_e2e_pipe_params_st *pipes, +int pipe_cnt, +int vlevel) +{ +int i, pipe_idx; +double dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; + +dc_assert_fp_enabled (); + +if (context->bw_ctx.dml.soc.min_dcfclk > dcfclk) + dcfclk = context->bw_ctx.dml.soc.min_dcfclk; + +/* We don't recalculate clocks for 0 pipe configs, which can block +* S0i3 as high clocks will block low power states +* Override any clocks that can block S0i3 to min here +*/ +if (pipe_cnt ==0 ) { + context->bw_ctx.bw.dcn.clk.dcfclk_khz = dcfclk;// always should be vlevel 0 +return ; + } + + pipes[0 ].clks_cfg.voltage = vlevel; + pipes[0 ].clks_cfg.dcfclk_mhz = dcfclk; + pipes[0 ].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; + +#if 0 // TODO +/* Set B: +* TODO +*/ +if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].valid) { +if (vlevel ==0 ) { + pipes[0 ].clks_cfg.voltage =1 ; + pipes[0 ].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[0 ].dcfclk_mhz; + } + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].dml_input.pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].dml_input.sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B ].dml_input.sr_exit_time_us; + } + context->bw_ctx.bw.dcn.watermarks.b.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.b.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + + pipes[0 ].clks_cfg.voltage = vlevel; + pipes[0 ].clks_cfg.dcfclk_mhz = dcfclk; + +/* Set C: +* TODO +*/ +if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].valid) { + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].dml_input.pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].dml_input.sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C ].dml_input.sr_exit_time_us; + } + context->bw_ctx.bw.dcn.watermarks.c.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.c.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + +/* Set D: +* TODO +*/ +if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].valid) { + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].dml_input.pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].dml_input.sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D ].dml_input.sr_exit_time_us; + } + context->bw_ctx.bw.dcn.watermarks.d.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.d.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; +#endif + +/* Set A: +* All clocks min required +* +* Set A calculated last so that following calculations are based on Set A +*/ + dc->res_pool->funcs->update_soc_for_wm_a (dc, context); + context->bw_ctx.bw.dcn.watermarks.a.urgent_ns =get_wm_urgent (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns =get_wm_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns =get_wm_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns =get_wm_dram_clock_change (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns =get_wm_z8_stutter_enter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_z8_ns =get_wm_z8_stutter_exit (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns =get_wm_memory_trip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom =get_fraction_of_urgent_bandwidth (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip =get_fraction_of_urgent_bandwidth_imm_flip (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; + context->bw_ctx.bw.dcn.watermarks.a.urgent_latency_ns =get_urgent_latency (&context->bw_ctx.dml, pipes, pipe_cnt) *1000 ; +/* TODO: remove: */ + context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; +/* end remove*/ + +for (i =0 , pipe_idx =0 ; i < dc->res_pool->pipe_count; i++) { +if (!context->res_ctx.pipe_ctx[i].stream) +continue ; + + pipes[pipe_idx].clks_cfg.dispclk_mhz =get_dispclk_calculated (&context->bw_ctx.dml, pipes, pipe_cnt); + pipes[pipe_idx].clks_cfg.dppclk_mhz =get_dppclk_calculated (&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx); + +if (dc->config.forced_clocks || dc->debug.max_disp_clk) { + pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0 ].dispclk_mhz; + pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0 ].dppclk_mhz; + } +if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz *1000 ) + pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz /1000.0 ; +if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz *1000 ) + pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz /1000.0 ; + + pipe_idx++; + } + +dcn20_calculate_dlg_params (dc, context, pipes, pipe_cnt, vlevel); +} + +void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params) +{ +struct clk_limit_table *clk_table = &bw_params->clk_table; +struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES ]; +unsigned int i, closest_clk_lvl; +int j; + +dc_assert_fp_enabled (); + +// Default clock levels are used for diags, which may lead to overclocking. +if (!IS_DIAG_DC (dc->ctx->dce_environment)) { +int max_dispclk_mhz =0 , max_dppclk_mhz =0 ; + + dcn3_1_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; + dcn3_1_ip.max_num_dpp = dc->res_pool->pipe_count; + dcn3_1_soc.num_chans = bw_params->num_channels; + +ASSERT (clk_table->num_entries); + +/* Prepass to find max clocks independent of voltage level. */ +for (i =0 ; i < clk_table->num_entries; ++i) { +if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) + max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; +if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) + max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; + } + +for (i =0 ; i < clk_table->num_entries; i++) { +/* loop backwards*/ +for (closest_clk_lvl =0 , j = dcn3_1_soc.num_states -1 ; j >=0 ; j--) { +if ((unsigned int ) dcn3_1_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { + closest_clk_lvl = j; +break ; + } + } + + clock_limits[i].state = i; + +/* Clocks dependent on voltage level. */ + clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; + clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; + clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; + clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz *2 * clk_table->entries[i].wck_ratio; + +/* Clocks independent of voltage level. */ + clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : + dcn3_1_soc.clock_limits[closest_clk_lvl].dispclk_mhz; + + clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : + dcn3_1_soc.clock_limits[closest_clk_lvl].dppclk_mhz; + + clock_limits[i].dram_bw_per_chan_gbps = dcn3_1_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; + clock_limits[i].dscclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dscclk_mhz; + clock_limits[i].dtbclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; + clock_limits[i].phyclk_d18_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; + clock_limits[i].phyclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_mhz; + } +for (i =0 ; i < clk_table->num_entries; i++) + dcn3_1_soc.clock_limits[i] = clock_limits[i]; +if (clk_table->num_entries) { + dcn3_1_soc.num_states = clk_table->num_entries; + } + } + + dcn3_1_soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz /1000.0 ; + dc->dml.soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz /1000.0 ; + +if (!IS_FPGA_MAXIMUS_DC (dc->ctx->dce_environment)) +dml_init_instance (&dc->dml, &dcn3_1_soc, &dcn3_1_ip,DML_PROJECT_DCN31 ); +else +dml_init_instance (&dc->dml, &dcn3_1_soc, &dcn3_1_ip,DML_PROJECT_DCN31_FPGA ); +}drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h new file +39
-0@@ -0,0 +1,39 @@ +/* +* Copyright 2019-2021 Advanced Micro Devices, Inc. +* +* Permission is hereby granted, free of charge, to any person obtaining a +* copy of this software and associated documentation files (the "Software"), +* to deal in the Software without restriction, including without limitation +* the rights to use, copy, modify, merge, publish, distribute, sublicense, +* and/or sell copies of the Software, and to permit persons to whom the +* Software is furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +* OTHER DEALINGS IN THE SOFTWARE. +* +* Authors: AMD +* +*/ + +#ifndef __DCN31_FPU_H__ +#define __DCN31_FPU_H__ + +#define DCN3_1_DEFAULT_DET_SIZE 384 + +void dcn31_calculate_wm_and_dlg_fp ( +struct dc *dc,struct dc_state *context, +display_e2e_pipe_params_st *pipes, +int pipe_cnt, +int vlevel); + +void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params); + +#endif /* __DCN31_FPU_H__*/
PATCH 12/13 drm/amd/display: move FPU related code from dcn315 to dml/dcn31 folder
In reply to Alex HungFrom: Melissa Wen <[email protected]>
Moves related structs and dcn315_update_bw_bounding_box from dcn315 driver code to dml/dcn31_fpu that centralizes FPU code for DCN 3.1x.
--- .../gpu/drm/amd/display/dc/dcn315/Makefile | 26 -- .../amd/display/dc/dcn315/dcn315_resource.c | 232 +----------------- .../amd/display/dc/dcn315/dcn315_resource.h | 3 + .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 228 +++++++++++++++++ .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 3 + 5 files changed, 235 insertions(+), 257 deletions(-)
drivers/gpu/drm/amd/display/dc/dcn315/Makefile +0 -26
@@ -25,32 +25,6 @@ DCN315 = dcn315_resource.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o += -msse2 -endif -endif - AMD_DAL_DCN315 = $(addprefix $(AMDDALPATH)/dc/dcn315/,$(DCN315)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN315)
drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c +1 -231
253 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.h +3 -0
@@ -31,6 +31,9 @@#define TO_DCN315_RES_POOL (pool)\ container_of(pool, struct dcn315_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_15_ip; +extern struct _vcs_dpi_ip_params_st dcn3_15_soc; +struct dcn315_resource_pool {struct resource_pool base; };
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c +228 -0
237 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h +3 -0
@@ -27,6 +27,8 @@#define __DCN31_FPU_H__#define DCN3_1_DEFAULT_DET_SIZE 384 +#define DCN3_15_DEFAULT_DET_SIZE 192 +#define DCN3_15_MIN_COMPBUF_SIZE_KB 128 void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context,
@@ -35,5 +37,6 @@ void dcn31_calculate_wm_and_dlg_fp( int vlevel);void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params); +void dcn315_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params);#endif /* __DCN31_FPU_H__*/
Signature · 1 line
2.35.1
PATCH 13/13 drm/amd/display: move FPU related code from dcn316 to dml/dcn31 folder
In reply to Alex HungFrom: Melissa Wen <[email protected]>
Moves FPU-related structs and dcn316_update_bw_bounding_box from dcn316 driver to dml/dcn31 that centralize FPU operations for DCN 3.1x
--- .../gpu/drm/amd/display/dc/dcn316/Makefile | 26 -- .../amd/display/dc/dcn316/dcn316_resource.c | 231 +----------------- .../amd/display/dc/dcn316/dcn316_resource.h | 3 + .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 229 +++++++++++++++++ .../drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 2 + 5 files changed, 235 insertions(+), 256 deletions(-)
drivers/gpu/drm/amd/display/dc/dcn316/Makefile +0 -26
@@ -25,32 +25,6 @@ DCN316 = dcn316_resource.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o += -msse2 -endif -endif - AMD_DAL_DCN316 = $(addprefix $(AMDDALPATH)/dc/dcn316/,$(DCN316)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN316)
drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c +1 -230
253 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.h +3 -0
@@ -31,6 +31,9 @@#define TO_DCN316_RES_POOL (pool)\ container_of(pool, struct dcn316_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_16_ip; +extern struct _vcs_dpi_ip_params_st dcn3_16_soc; +struct dcn316_resource_pool {struct resource_pool base; };
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c +229 -0
238 lines not on this page · raw message
drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h +2 -0
@@ -29,6 +29,7 @@#define DCN3_1_DEFAULT_DET_SIZE 384#define DCN3_15_DEFAULT_DET_SIZE 192#define DCN3_15_MIN_COMPBUF_SIZE_KB 128 +#define DCN3_16_DEFAULT_DET_SIZE 192 void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context,
@@ -38,5 +39,6 @@ void dcn31_calculate_wm_and_dlg_fp(void dcn31_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params);void dcn315_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params); +void dcn316_update_bw_bounding_box (struct dc *dc,struct clk_bw_params *bw_params);#endif /* __DCN31_FPU_H__*/
Signature · 1 line
2.35.1